- Email: [email protected]
Paroxetine Treatment of Primary Insomnia in Older Adults
BRIEF REPORT Paroxetine Treatment of Primary Insomnia in Older Adults Charles F. Reynolds III, M.D. Daniel J. Buysse, M.D. Mark D. Miller, M.D. Bruce G. Pollock, M.D., Ph.D. Martica Hall, Ph.D. Sati Mazumdar, Ph.D.
Objective: The goal of this study was to test the efficacy of paroxetine for primary insomnia in older adults. Methods: Adults over age 55 with primary insomnia were randomly assigned to six weeks of double-blind treatment with paroxetine (N ⫽14) or placebo (N⫽13). Outcome measures included persistence or resolution of insomnia; weekly diary measures of sleep quality, daytime alertness, and mood; and pre-/postpolysomnographic measures of sleep latency, wake time after sleep onset, and sleep efficiency. Results: Although weekly diary-based measures showed improved subjective sleep quality, daytime alertness, and mood with paroxetine, the authors observed no effect on sleep efficiency or categorical response rates. Conclusion: Paroxetine appears to be ineffective for treating primary insomnia in old age. (Am J Geriatr Psychiatry 2006; 14:803–807) Key Words: Primary insomnia, antidepressant, paroxetine, old age
C
hronic insomnia has a prevalence of 5%–10% in community samples and is associated with elevated levels of anxiety, depression, distress, and general medical burden, as well as with interpersonal and occupational impairments.1,2 Chronic insomnia
that extends over a one-year period is a risk factor for the development of major depression.1 Untreated chronic insomnia is most highly prevalent and persistent in the elderly.2 Because use of sedative hypnotics in the elderly raises safety concerns regarding cognitive impairment, falls, and sleep apnea,3 clinicians often use antidepressants as hypnotics despite the paucity of empiric support for this practice. Some clinical trials have shown evidence for the efficacy of sedating antidepressants such as trazodone, doxepin, and trimipramine,4 – 6 but the cumulative experience with these interventions remains limited.7 At the same time, studies of nonsedating antidepressants, including selective serotonin reuptake inhibitors, in patients whose insomnia is associated with a mood disorder, suggest that mechanisms other than sedation may lead to improvement in self-reported sleep quality.8 Given the relationship between chronic insomnia and depression, we chose to examine the effectiveness of paroxetine for primary insomnia (as defined by the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition [DSM-IV]). In an open pilot study of paroxetine administered over six weeks in 15 patients ranging in age from 35–75,9 we reported that seven of 14 participants no longer met diagnostic criteria for primary insomnia after 6 weeks of paroxetine pharmacotherapy (median dose: 20 mg; range: 5–30 mg) combined with instruction in good sleep hygiene. Self-reported sleep quality and mood improved with treatment, but the quantity of sleep as measured polysomnographically and by diary did not change. We hypothesized that paroxetine at bedtime, in combination with sleep hygiene education, would produce a higher rate of response to acute treatment than placebo. We also predicted greater improvement in polysomnographic measures of sleep latency, wakefulness after sleep onset, and sleep efficiency; and in daily diary-based measures of subjective sleep quality, alertness, and mood.
Received December 19, 2005; revised February 1, 2006; accepted February 11, 2006. From the Advanced Center for Interventions and Services Research for Late-Life Mood Disorders (ACISR/LLMD) and the John A. Hartford Center of Excellence in Geriatric Psychiatry, Western Psychiatric Institute and Clinic, Pittsburgh, PA (CFR, DJB, MDM, MH); Division of Geriatric Psychiatry, University of Toronto, Faculty of Medicine, The Rotman Research Institute, Baycrest Centre for Geriatric Care, Toronto, Ontario, Canada (BGP); and Department of Biostatistics, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, PA (SM). Send correspondence and reprint requests to Dr. Charles F. Reynolds III, Western Psychiatric Institute and Clinic, 3811 O’Hara St., Pittsburgh, PA. e-mail: [email protected] © 2006 American Association for Geriatric Psychiatry
Am J Geriatr Psychiatry 14:9, September 2006
803
Paroxetine Treatment of Primary Insomnia
SUBJECTS AND METHODS Patients aged 55 and older were eligible for participation if they met DSM–IV criteria for primary insomnia.10 These criteria include: 1) difficulty initiating or maintaining sleep, or nonrestorative sleep, for at least one month; 2) clinically significant distress or functional impairment; 3) the sleep disturbance does not occur exclusively during the course of narcolepsy, breathing-related sleep disorder, circadian rhythm sleep disorder, or a parasomnia; 4) the disturbance does not occur exclusively during the course of another mental disorder; and 5) the disturbance is not the result of the direct physiological effects of a substance or a general medical condition. Clinical assessment by a study investigator and the project coordinator used structured sleep and psychiatric interviewing (SCID for DSM–IV) to determine diagnosis. Patients spent two consecutive nights in the WPIC Clinical Neuroscience Center to obtain screening polysomnography for the purpose of ruling out sleep-disordered breathing and periodic limb movements as well as to obtain a second baseline night. Urine toxicology before the start of treatment and repeated at the end of the trial was also obtained to rule out self-medication with benzodiazepines and other substances. The two treatment groups (paroxetine, placebo) did not differ significantly on any demographic or clinical measures. Mean age of paroxetine-treated subjects was 67.4 (standard deviation: 10.5) versus 66.5 (standard deviation: 7.4) in placebo-treated subjects. We screened 265 volunteers recruited by clinical referral and media announcements. Two hundred nine failed the screen and of the remaining 56 who signed informed consent, 27 started treatment. The overwhelming reason for screen failure was not meeting DSM–IV criteria for primary insomnia, i.e., most patients had chronic insomnia better accounted for by coexisting medical, psychiatric, or substance abuse disorders. Among the 29 patients signing informed consent but not starting treatment, we excluded five because of apnea– hypopnea indices of greater than 20; nine because of periodic limb movement disorders with periodic limb movement arousal indices of greater than 15, three did not have sufficiently disturbed sleep on the Pittsburgh Sleep Quality Index (8 or higher), one had sinus bradycar-
804
dia (representing a relative contraindication to paroxetine), nine chose to drop out before treatment because of respondent burden, one was taking a medication that contraindicated selective serotonin reuptake inhibitor pharmacotherapy, and one was facing surgery. On balance, approximately one in 10 patients screened joined treatment. All participants provided written informed consent. Patients were randomly assigned to receive paroxetine or placebo, initiated as a 10-mg dose and adjusted after two weeks based on the presence of possible side effects up to a maximum dose of 20 mg daily. Treatment and dose adjustments were conducted under double-blind conditions. For the 14 subjects on paroxetine, six had a final dose of 10 mg and eight had a final dose of 20 mg (mean: 15.7 [standard deviation: 5.1]). Seven of 27 subjects dropped out before completing treatment, four of 14 on paroxetine (failure to improve, two; rash, one; daytime use of stimulant, one) and 3 of 13 on placebo (two failure to improve, one respondent burden). Four paroxetine-treated subjects dropped out by weeks 3 and 4; three placebo-treated subjects dropped out by four weeks. After six weeks of double-blind treatment, patients were assessed for diagnostic response status by a blinded evaluator (DJB) with Pittsburgh Sleep Diarybased measures of sleep quality, daytime alertness, and mood; and polysomnographic sleep measures comprising sleep latency (time to fall asleep), wakefulness after sleep onset, and overall sleep efficiency. To test the hypothesis of differential improvement in participants randomized to paroxetine, we performed repeated-measures mixed-model analyses of the seven outcome measures to test group, time, and group ⫻ time interactions. These analyses did not impute missing data from subject dropouts; we allowed the mixed model analysis to solve without imputation. We performed a pre-/postanalysis of polysomnographic measures (sleep latency, wakefulness after sleep onset, sleep efficiency). To test diagnostic response rate, we used a chi-squared test. The analyses were done on transformed data so that outliers seen in the figure were much reduced. For the variable “mood,” we used the assumption of compound symmetry, which gave reasonable estimates of the correlation in the data and produced a different df value.
Am J Geriatr Psychiatry 14:9, September 2006
Reynolds et al.
RESULTS Diagnostic Response Status Seven of 14 paroxetine-treated patients improved and no longer met diagnostic criteria for primary insomnia at the end of the trial versus 5 of 13 placebo-treated patients (chi squared⫽3.4, df⫽1, p ⫽ 0.19). Polysomnographic Measures Compared (Figure 1) with placebo, patients taking paroxetine demonstrated increased time to fall asleep (F⫽4.61, df⫽1,25, p⫽0.04) but decreased wakefulness after sleep onset (F⫽6.09, df⫽1,25, p⫽ 0.02) (significant group ⫻ time interaction for both variables). However, the groups did not show differential change in overall sleep efficiency (a ratio of the time spent asleep divided by total recording period). Diary-Based Measures Paroxetine-treated patients reported more improvement (group ⫻ time interaction) than placebotreated patients in diary-based measures of subjective sleep quality (F⫽5.42, df⫽1,25, p ⫽0.03), daytime alertness (F⫽8.19, df ⫽1,25, p ⫽0.008), and mood (F⫽6.12, df⫽1,136, p ⫽0.02). Sleep quality demonstrated a pre- to postmean change of 16.4 for paroxetine versus 4.4 for placebo (Cohen’s d for interaction ⫽0.93). Daytime alertness improved to a greater extent with paroxetine than placebo (15.6 versus 0.6; Cohen’s d for interaction⫽1.14) as did mood/well-being (13.0 versus 2.6; Cohen’s d⫽0.42). For each measure, the inflection point signaling the start of greater improvement was apparent at 3 weeks. The consistently greater improvement in diary-based measures of sleep quality, mood, and daytime alertness is evident from the narrower distribution of final 6-week scores among paroxetine- compared with placebo-treated subjects (Figure 1).
DISCUSSION Paroxetine combined with sleep hygiene education demonstrated only a small effect, as compared with
Am J Geriatr Psychiatry 14:9, September 2006
placebo ⫹ sleep hygiene, on overall categorical response rate, 50% versus 38.5%. The 50% response rate for paroxetine ⫹ sleep hygiene was identical to the 50% response rate reported in our open pilot study.9 We did detect differentially greater improvement for paroxetine in diary-based measures of subjective sleep quality, daytime alertness, and mood, but quantity and efficiency of sleep did not change. A similar pattern of improvement in diary-based measures of sleep quality and mood was seen in our earlier open-trial study of paroxetine. Diary-based measures of sleep quality, mood, and alertness all demonstrated a sharp upward deflection (i.e., start of improvement) at week 3 of treatment and sustained an upward trajectory until the end of the 6-week trial (Figure 1). The small study group is an important limitation in our ability to interpret the findings. We were surprised by the high screen-to-admit ratio of 10 to 1, reflecting the fact that most of the help-seeking elderly people with persistent insomnia complaints did not satisfy DSM–IV criteria for primary insomnia but had, instead, coexisting medical and psychiatric diagnoses that better accounted for their sleep complaint. This raises the issue of whether DSM–IV primary insomnia has clinical relevance to geriatric practice. We suggest that a more clinically relevant approach would be to study patients with persistent insomnia complaints without imposing DSM–IV medical or psychiatric exclusion criteria. The recent National Institutes of Health State of the Science Conference on the Manifestations and Morbidity of Chronic Insomnia7 reached similar conclusions. The conference panel recommended that studies begin to examine individuals with “comorbid insomnia” given the difficulty of determining what actually constitutes “secondary” insomnia. Another factor of great relevance to insomnia management in the elderly is the inevitable coexistence of multiple medical problems with high numbers of coprescribed and over-the-counter medications. The small benefit evidenced with paroxetine raises an important issue of the appropriateness of adding another medication to what is likely to be an already complex and expensive regimen. It is arguably more appropriate to test the value of brief behavioral treatments for insomnia in medically heterogeneous primary care patients, in whom benefit may be realized without the cost or potential hazard of adding another medication.7
805
Paroxetine Treatment of Primary Insomnia
FIGURE 1.
Paroxetine Effects on Sleep, Daytime Alertness, and Mood: Group and Individual Data
Note: We detected no consistent effect of paroxetine on polysomnogrphic (PSG) measures but did see consistent effects in improvements of diary-based sleep quality, daytime alertness, and mood.
806
Am J Geriatr Psychiatry 14:9, September 2006
Reynolds et al. In summary, these data do not support the use of paroxetine in treating chronic primary insomnia in old age. The existing literature more strongly supports the use of behavioral strategies such as stimulus control and sleep restriction.7 In the spirit of bridging science and service, the challenge posed is to find ways of disseminating practical versions of behavioral treatments for use by the elderly in primary care. If these treatments can be successfully implemented in primary care, then the next challenge will be to test their value in preventing the complications of insomnia in later life, including falls and depression.
Supported in part by P30 MH071944, P30 MH52247, R01 MH37869, MH65416, AG20677, RR00056, MH24652, AG00972; the John A. Hartford Foundation Center of Excellence in Geriatric Psychiatry, and the University of Pittsburgh Medical Center endowed professorship in geriatric psychiatry. We thank the staff of the Clinical Trials Management Unit of the ACISR/LLMD for their recruitment, assessment, and care of the participants in this trial, and the staff of the Data Management/Analysis unit for the analysis of the data reported. GlaxoSmithKline provided paroxetine for the conduct of this study, and had no other role in the conduct of the study or data analysis.
References 1. Ford DE, Kamerow DB: Epidemiologic study of sleep disturbances and psychiatric disorders. An opportunity for prevention? JAMA 1989; 262:1479 –1484 2. Ganguli M, Reynolds CF, Gilby JE: Prevalence and persistence of sleep complaints in a rural elderly community sample: the MoVIES Project. J Am Geriatr Soc 1996; 44:778 –784 3. Nowell PD, Mazumdar S, Buysse DJ, et al: Benzodiazepines and zolpidem for chronic insomnia: a meta-analysis of treatment efficacy. JAMA 1997; 278:2170 –2177 4. Walsh JK, Erman M, Erwin CW, et al: Subjective hypnotic efficacy of trazodone and zolpidem in DSMIII-R primary insomnia. Human Psychopharmacology 1998; 13:191–198 5. Hohagen F, Montero RF, Weiss E, et al: Treatment of primary insomnia with trimipramine: an alternative to benzodiazepine hypnotics? Eur Arch Psychiatry Clin Neurosci 1994; 244:65–72 6. Hajak G, Rodenbeck A, Voderholzer U, et al: Doxepin in the
Am J Geriatr Psychiatry 14:9, September 2006
treatment of primary insomnia: a placebo-controlled, doubleblind, polysomnographic study. J Clin Psychiatry 2001; 62:453– 463 7. NIH state-of-the-science conference statement on manifestations and management of chronic insomnia in adults. Bethesda, MD, National Institutes of Health, 2005. 8. Armitage R, Yonkers K, Cole D, et al: A multi-center, double-blind comparison of the effects of nefazodone and fluoxetine on sleep architecture and quality of sleep in depressed outpatients. J Clin Psychopharmacol 1997; 17:161–168 9. Nowell PD, Reynolds CF, Buysse DJ, et al: Paroxetine in the treatment of primary insomnia: preliminary clinical and EEG sleep data. J Clin Psychiatry 1999; 60:89 –95 10. First M, Spitzer RL, Gibbon M, et al: Structured Clinical Interview for DSM–IV Axis I Disorders–Patient Edition (SCID-I/P). New York, State Psychiatric Institute, 1995
807
Objective: The goal of this study was to test the efficacy of paroxetine for primary insomnia in older adults. Methods: Adults over age 55 with primary insomnia were randomly assigned to six weeks of double-blind treatment with paroxetine (N ⫽14) or placebo (N⫽13). Outcome measures included persistence or resolution of insomnia; weekly diary measures of sleep quality, daytime alertness, and mood; and pre-/postpolysomnographic measures of sleep latency, wake time after sleep onset, and sleep efficiency. Results: Although weekly diary-based measures showed improved subjective sleep quality, daytime alertness, and mood with paroxetine, the authors observed no effect on sleep efficiency or categorical response rates. Conclusion: Paroxetine appears to be ineffective for treating primary insomnia in old age. (Am J Geriatr Psychiatry 2006; 14:803–807) Key Words: Primary insomnia, antidepressant, paroxetine, old age
C
hronic insomnia has a prevalence of 5%–10% in community samples and is associated with elevated levels of anxiety, depression, distress, and general medical burden, as well as with interpersonal and occupational impairments.1,2 Chronic insomnia
that extends over a one-year period is a risk factor for the development of major depression.1 Untreated chronic insomnia is most highly prevalent and persistent in the elderly.2 Because use of sedative hypnotics in the elderly raises safety concerns regarding cognitive impairment, falls, and sleep apnea,3 clinicians often use antidepressants as hypnotics despite the paucity of empiric support for this practice. Some clinical trials have shown evidence for the efficacy of sedating antidepressants such as trazodone, doxepin, and trimipramine,4 – 6 but the cumulative experience with these interventions remains limited.7 At the same time, studies of nonsedating antidepressants, including selective serotonin reuptake inhibitors, in patients whose insomnia is associated with a mood disorder, suggest that mechanisms other than sedation may lead to improvement in self-reported sleep quality.8 Given the relationship between chronic insomnia and depression, we chose to examine the effectiveness of paroxetine for primary insomnia (as defined by the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition [DSM-IV]). In an open pilot study of paroxetine administered over six weeks in 15 patients ranging in age from 35–75,9 we reported that seven of 14 participants no longer met diagnostic criteria for primary insomnia after 6 weeks of paroxetine pharmacotherapy (median dose: 20 mg; range: 5–30 mg) combined with instruction in good sleep hygiene. Self-reported sleep quality and mood improved with treatment, but the quantity of sleep as measured polysomnographically and by diary did not change. We hypothesized that paroxetine at bedtime, in combination with sleep hygiene education, would produce a higher rate of response to acute treatment than placebo. We also predicted greater improvement in polysomnographic measures of sleep latency, wakefulness after sleep onset, and sleep efficiency; and in daily diary-based measures of subjective sleep quality, alertness, and mood.
Received December 19, 2005; revised February 1, 2006; accepted February 11, 2006. From the Advanced Center for Interventions and Services Research for Late-Life Mood Disorders (ACISR/LLMD) and the John A. Hartford Center of Excellence in Geriatric Psychiatry, Western Psychiatric Institute and Clinic, Pittsburgh, PA (CFR, DJB, MDM, MH); Division of Geriatric Psychiatry, University of Toronto, Faculty of Medicine, The Rotman Research Institute, Baycrest Centre for Geriatric Care, Toronto, Ontario, Canada (BGP); and Department of Biostatistics, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, PA (SM). Send correspondence and reprint requests to Dr. Charles F. Reynolds III, Western Psychiatric Institute and Clinic, 3811 O’Hara St., Pittsburgh, PA. e-mail: [email protected] © 2006 American Association for Geriatric Psychiatry
Am J Geriatr Psychiatry 14:9, September 2006
803
Paroxetine Treatment of Primary Insomnia
SUBJECTS AND METHODS Patients aged 55 and older were eligible for participation if they met DSM–IV criteria for primary insomnia.10 These criteria include: 1) difficulty initiating or maintaining sleep, or nonrestorative sleep, for at least one month; 2) clinically significant distress or functional impairment; 3) the sleep disturbance does not occur exclusively during the course of narcolepsy, breathing-related sleep disorder, circadian rhythm sleep disorder, or a parasomnia; 4) the disturbance does not occur exclusively during the course of another mental disorder; and 5) the disturbance is not the result of the direct physiological effects of a substance or a general medical condition. Clinical assessment by a study investigator and the project coordinator used structured sleep and psychiatric interviewing (SCID for DSM–IV) to determine diagnosis. Patients spent two consecutive nights in the WPIC Clinical Neuroscience Center to obtain screening polysomnography for the purpose of ruling out sleep-disordered breathing and periodic limb movements as well as to obtain a second baseline night. Urine toxicology before the start of treatment and repeated at the end of the trial was also obtained to rule out self-medication with benzodiazepines and other substances. The two treatment groups (paroxetine, placebo) did not differ significantly on any demographic or clinical measures. Mean age of paroxetine-treated subjects was 67.4 (standard deviation: 10.5) versus 66.5 (standard deviation: 7.4) in placebo-treated subjects. We screened 265 volunteers recruited by clinical referral and media announcements. Two hundred nine failed the screen and of the remaining 56 who signed informed consent, 27 started treatment. The overwhelming reason for screen failure was not meeting DSM–IV criteria for primary insomnia, i.e., most patients had chronic insomnia better accounted for by coexisting medical, psychiatric, or substance abuse disorders. Among the 29 patients signing informed consent but not starting treatment, we excluded five because of apnea– hypopnea indices of greater than 20; nine because of periodic limb movement disorders with periodic limb movement arousal indices of greater than 15, three did not have sufficiently disturbed sleep on the Pittsburgh Sleep Quality Index (8 or higher), one had sinus bradycar-
804
dia (representing a relative contraindication to paroxetine), nine chose to drop out before treatment because of respondent burden, one was taking a medication that contraindicated selective serotonin reuptake inhibitor pharmacotherapy, and one was facing surgery. On balance, approximately one in 10 patients screened joined treatment. All participants provided written informed consent. Patients were randomly assigned to receive paroxetine or placebo, initiated as a 10-mg dose and adjusted after two weeks based on the presence of possible side effects up to a maximum dose of 20 mg daily. Treatment and dose adjustments were conducted under double-blind conditions. For the 14 subjects on paroxetine, six had a final dose of 10 mg and eight had a final dose of 20 mg (mean: 15.7 [standard deviation: 5.1]). Seven of 27 subjects dropped out before completing treatment, four of 14 on paroxetine (failure to improve, two; rash, one; daytime use of stimulant, one) and 3 of 13 on placebo (two failure to improve, one respondent burden). Four paroxetine-treated subjects dropped out by weeks 3 and 4; three placebo-treated subjects dropped out by four weeks. After six weeks of double-blind treatment, patients were assessed for diagnostic response status by a blinded evaluator (DJB) with Pittsburgh Sleep Diarybased measures of sleep quality, daytime alertness, and mood; and polysomnographic sleep measures comprising sleep latency (time to fall asleep), wakefulness after sleep onset, and overall sleep efficiency. To test the hypothesis of differential improvement in participants randomized to paroxetine, we performed repeated-measures mixed-model analyses of the seven outcome measures to test group, time, and group ⫻ time interactions. These analyses did not impute missing data from subject dropouts; we allowed the mixed model analysis to solve without imputation. We performed a pre-/postanalysis of polysomnographic measures (sleep latency, wakefulness after sleep onset, sleep efficiency). To test diagnostic response rate, we used a chi-squared test. The analyses were done on transformed data so that outliers seen in the figure were much reduced. For the variable “mood,” we used the assumption of compound symmetry, which gave reasonable estimates of the correlation in the data and produced a different df value.
Am J Geriatr Psychiatry 14:9, September 2006
Reynolds et al.
RESULTS Diagnostic Response Status Seven of 14 paroxetine-treated patients improved and no longer met diagnostic criteria for primary insomnia at the end of the trial versus 5 of 13 placebo-treated patients (chi squared⫽3.4, df⫽1, p ⫽ 0.19). Polysomnographic Measures Compared (Figure 1) with placebo, patients taking paroxetine demonstrated increased time to fall asleep (F⫽4.61, df⫽1,25, p⫽0.04) but decreased wakefulness after sleep onset (F⫽6.09, df⫽1,25, p⫽ 0.02) (significant group ⫻ time interaction for both variables). However, the groups did not show differential change in overall sleep efficiency (a ratio of the time spent asleep divided by total recording period). Diary-Based Measures Paroxetine-treated patients reported more improvement (group ⫻ time interaction) than placebotreated patients in diary-based measures of subjective sleep quality (F⫽5.42, df⫽1,25, p ⫽0.03), daytime alertness (F⫽8.19, df ⫽1,25, p ⫽0.008), and mood (F⫽6.12, df⫽1,136, p ⫽0.02). Sleep quality demonstrated a pre- to postmean change of 16.4 for paroxetine versus 4.4 for placebo (Cohen’s d for interaction ⫽0.93). Daytime alertness improved to a greater extent with paroxetine than placebo (15.6 versus 0.6; Cohen’s d for interaction⫽1.14) as did mood/well-being (13.0 versus 2.6; Cohen’s d⫽0.42). For each measure, the inflection point signaling the start of greater improvement was apparent at 3 weeks. The consistently greater improvement in diary-based measures of sleep quality, mood, and daytime alertness is evident from the narrower distribution of final 6-week scores among paroxetine- compared with placebo-treated subjects (Figure 1).
DISCUSSION Paroxetine combined with sleep hygiene education demonstrated only a small effect, as compared with
Am J Geriatr Psychiatry 14:9, September 2006
placebo ⫹ sleep hygiene, on overall categorical response rate, 50% versus 38.5%. The 50% response rate for paroxetine ⫹ sleep hygiene was identical to the 50% response rate reported in our open pilot study.9 We did detect differentially greater improvement for paroxetine in diary-based measures of subjective sleep quality, daytime alertness, and mood, but quantity and efficiency of sleep did not change. A similar pattern of improvement in diary-based measures of sleep quality and mood was seen in our earlier open-trial study of paroxetine. Diary-based measures of sleep quality, mood, and alertness all demonstrated a sharp upward deflection (i.e., start of improvement) at week 3 of treatment and sustained an upward trajectory until the end of the 6-week trial (Figure 1). The small study group is an important limitation in our ability to interpret the findings. We were surprised by the high screen-to-admit ratio of 10 to 1, reflecting the fact that most of the help-seeking elderly people with persistent insomnia complaints did not satisfy DSM–IV criteria for primary insomnia but had, instead, coexisting medical and psychiatric diagnoses that better accounted for their sleep complaint. This raises the issue of whether DSM–IV primary insomnia has clinical relevance to geriatric practice. We suggest that a more clinically relevant approach would be to study patients with persistent insomnia complaints without imposing DSM–IV medical or psychiatric exclusion criteria. The recent National Institutes of Health State of the Science Conference on the Manifestations and Morbidity of Chronic Insomnia7 reached similar conclusions. The conference panel recommended that studies begin to examine individuals with “comorbid insomnia” given the difficulty of determining what actually constitutes “secondary” insomnia. Another factor of great relevance to insomnia management in the elderly is the inevitable coexistence of multiple medical problems with high numbers of coprescribed and over-the-counter medications. The small benefit evidenced with paroxetine raises an important issue of the appropriateness of adding another medication to what is likely to be an already complex and expensive regimen. It is arguably more appropriate to test the value of brief behavioral treatments for insomnia in medically heterogeneous primary care patients, in whom benefit may be realized without the cost or potential hazard of adding another medication.7
805
Paroxetine Treatment of Primary Insomnia
FIGURE 1.
Paroxetine Effects on Sleep, Daytime Alertness, and Mood: Group and Individual Data
Note: We detected no consistent effect of paroxetine on polysomnogrphic (PSG) measures but did see consistent effects in improvements of diary-based sleep quality, daytime alertness, and mood.
806
Am J Geriatr Psychiatry 14:9, September 2006
Reynolds et al. In summary, these data do not support the use of paroxetine in treating chronic primary insomnia in old age. The existing literature more strongly supports the use of behavioral strategies such as stimulus control and sleep restriction.7 In the spirit of bridging science and service, the challenge posed is to find ways of disseminating practical versions of behavioral treatments for use by the elderly in primary care. If these treatments can be successfully implemented in primary care, then the next challenge will be to test their value in preventing the complications of insomnia in later life, including falls and depression.
Supported in part by P30 MH071944, P30 MH52247, R01 MH37869, MH65416, AG20677, RR00056, MH24652, AG00972; the John A. Hartford Foundation Center of Excellence in Geriatric Psychiatry, and the University of Pittsburgh Medical Center endowed professorship in geriatric psychiatry. We thank the staff of the Clinical Trials Management Unit of the ACISR/LLMD for their recruitment, assessment, and care of the participants in this trial, and the staff of the Data Management/Analysis unit for the analysis of the data reported. GlaxoSmithKline provided paroxetine for the conduct of this study, and had no other role in the conduct of the study or data analysis.
References 1. Ford DE, Kamerow DB: Epidemiologic study of sleep disturbances and psychiatric disorders. An opportunity for prevention? JAMA 1989; 262:1479 –1484 2. Ganguli M, Reynolds CF, Gilby JE: Prevalence and persistence of sleep complaints in a rural elderly community sample: the MoVIES Project. J Am Geriatr Soc 1996; 44:778 –784 3. Nowell PD, Mazumdar S, Buysse DJ, et al: Benzodiazepines and zolpidem for chronic insomnia: a meta-analysis of treatment efficacy. JAMA 1997; 278:2170 –2177 4. Walsh JK, Erman M, Erwin CW, et al: Subjective hypnotic efficacy of trazodone and zolpidem in DSMIII-R primary insomnia. Human Psychopharmacology 1998; 13:191–198 5. Hohagen F, Montero RF, Weiss E, et al: Treatment of primary insomnia with trimipramine: an alternative to benzodiazepine hypnotics? Eur Arch Psychiatry Clin Neurosci 1994; 244:65–72 6. Hajak G, Rodenbeck A, Voderholzer U, et al: Doxepin in the
Am J Geriatr Psychiatry 14:9, September 2006
treatment of primary insomnia: a placebo-controlled, doubleblind, polysomnographic study. J Clin Psychiatry 2001; 62:453– 463 7. NIH state-of-the-science conference statement on manifestations and management of chronic insomnia in adults. Bethesda, MD, National Institutes of Health, 2005. 8. Armitage R, Yonkers K, Cole D, et al: A multi-center, double-blind comparison of the effects of nefazodone and fluoxetine on sleep architecture and quality of sleep in depressed outpatients. J Clin Psychopharmacol 1997; 17:161–168 9. Nowell PD, Reynolds CF, Buysse DJ, et al: Paroxetine in the treatment of primary insomnia: preliminary clinical and EEG sleep data. J Clin Psychiatry 1999; 60:89 –95 10. First M, Spitzer RL, Gibbon M, et al: Structured Clinical Interview for DSM–IV Axis I Disorders–Patient Edition (SCID-I/P). New York, State Psychiatric Institute, 1995
807