To identify whether baseline demographic factors or subjective sleep variables are associated with the outcomes following treatment with eszopiclone using data from a recent randomized controlled trial of 78 Japanese subjects with insomnia who were treated with 2 mg eszopiclone per day.
We performed a post hoc analysis of factors including sleep latency (SL), wake time after sleep onset (WASO) (both assessed via sleep diaries), and several demographic variables. Subjects with a SL or WASO > 30 min at baseline and with evaluable SL/WASO data at Week 4 were included in SL and WASO remitter analyses, respectively; those with a SL or WASO ≤ 30 min at Week 4 were defined as SL or WASO remitters, respectively. Threshold baseline SL and WASO values for identification of remitters were determined.
No relationships between subjectively assessed therapeutic outcomes and demographic factors were identified. Patients with shorter SL and lower WASO values at baseline showed better outcomes following treatment with eszopiclone in terms of SL and WASO changes, respectively. Baseline SL of 75 min and baseline WASO of 80 min were selected as arbitrary cutoff values for determination of SL and WASO remitters/non-remitters, respectively.
These findings may help clinicians to predict their patients' outcomes in response to standard doses of eszopiclone in clinical practice.
Inoue Y, Kamijo A, Nagai R. Patient background factors affecting the therapeutic outcomes in response to eszopiclone in adult patients with chronic insomnia: a post hoc analysis of a double-blind phase III study in Japan. J Clin Sleep Med 2015;11(10):1171–1178.
Insomnia is a highly prevalent condition: nocturnal insomnia symptoms are reported by approximately 30% of adults, with 6% to 10% of individuals meeting the diagnostic criteria for an insomnia disorder.1,2 Nocturnal insomnia consists of difficulty initiating sleep (sleep-onset insomnia) and difficulty maintaining sleep (sleep maintenance insomnia); wake after sleep onset is regarded as the main component of the latter.2,3 Insomnia can lead to the development of depression4 and decreased quality of life.5 In the elderly, insomnia is more common in individuals with a greater number of comorbidities, suggesting that the disorder is likely to occur in association with these comorbidities.6,7 Among the pharmacological treatments for insomnia, sedating hypnotic agents such as benzodiazepine receptor agonists with short-eliminating half-lives are widely accepted as the first-line treatment.8 In drug development studies, eszopiclone, a benzodiazepine receptor agonist, has shown good efficacy for treating chronic insomnia, providing a significant reduction in sleep latency (SL), increased total sleep time (TST), and reduced wake time after sleep onset (WASO).9–12 However, in clinical settings, certain individuals fail to respond sufficiently to hypnotic agents, leading to concerns about the use of ineffective medications and the risk of adverse drug reactions associated with long-term treatment or dose increases, including parasomnia,13 tolerance, and dependence.14 In such individuals, cognitive behavioral therapy for insomnia or related approaches can be used as an alternative to or in addition to sedating hypnotics.8 To promote the appropriate use of hypnotics, it would be desirable to be able to predict a patient's response to hypnotics prior to the start of treatment. However, most of the earlier clinical studies on hypnotics in patients with insomnia evaluated the changes in efficacy measures from baseline, using placebo as a control.9–12,15–18 There has been no report to date on the factors associated with the response to pharmacological treatment in patients with insomnia. In any analysis aimed at identifying potential responders, it is necessary to examine the effects of treatment on both sleep-onset insomnia (for which a response would be indicated by shortened SL) and sleep maintenance insomnia (for which a decrease in WASO would indicate a response).
Current Knowledge/Study Rationale: Pharmacological agents used to treat insomnia, such as eszopiclone, improve sleep-related assessments in most patients, but some patients show inadequate or no improvements in sleep. Therefore, it is important to identify which clinical factors are associated with the outcomes in response to eszopiclone to help clinicians predict their patients' responses to hypnotics before starting treatment.
Study Impact: Patients with shorter sleep latency (SL) and lower wake time after sleep onset (WASO) at baseline showed better outcomes following treatment with eszopiclone in terms of the changes in SL and WASO. The present results suggest that 2 mg eszopiclone per day is effective for treating patients with moderate sleep-onset insomnia with SL < 75 min or patients with sleep maintenance insomnia with WASO < 80 min.
Eszopiclone is a member of the nonbenzodiazepine hypnotic class of drugs. It acts as an allosteric modulator of the gamma-aminobutyric acid type A receptor with a half-life of approximately 6 hours.19 It was reported that eszopiclone improves both sleep onset and sleep maintenance.12 The safety and efficacy of eszopiclone were recently demonstrated in both elderly and nonelderly Japanese patients.20 That study was a two-phase study consisting of an initial treatment period of 4 weeks followed by a second treatment period from Week 4 to Week 24. Patients were randomized to different dose groups (1, 2, or 3 mg eszopiclone per day), but the trial was not placebo-controlled. In that study, the relationships between the clinical characteristics of patients and their responses to treatment were not investigated. Therefore, in the present study, we performed a post hoc analysis of factors that might affect the outcomes in response to treatment with eszopiclone using data from the initial treatment period of the randomized controlled trial during which the patients kept sleep diaries.20 Considering that there are age21 and gender22 differences in the prevalence of insomnia, our hypothesis was that age, gender, and other demographic or clinical variables would be associated with the outcomes following treatment with eszopiclone. We also determined the cutoff values for baseline sleep variables prior to starting treatment that could help identify remitters following eszopiclone treatment.
We used data from a multicenter, 24-week, randomized, double-blind, parallel-group study of eszopiclone in Japanese subjects with chronic insomnia.20 The protocol for that study was approved by the institutional review boards at each of the 46 sites involved. All patients provided written informed consent for participation in the study. The trial was conducted in accordance with the principles of the Declaration of Helsinki and Japan Good Clinical Practice, and registered at ClinicalTrials.gov (#NCT00770692).
The inclusion and exclusion criteria are described in more detail elsewhere,20 but, in brief, included the following. Inclusion criteria: age 20–84 years; outpatients with a diagnosis of primary insomnia according to the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR)23; and SL ≥ 30 min on ≥ 3 nights (including ≥ 2 consecutive nights) and TST ≤ 390 min on ≥ 3 nights (including ≥ 2 consecutive nights) as recorded in the patients' sleep diaries during the screening period. The criterion of SL ≥ 30 min on ≥ 3 nights was selected in accordance with a previous study.24 No WASO severity criterion was required to be satisfied for entry into the study. Exclusion criteria included the following: behavioral disorders (see Uchimura et al.20 for details); drug-induced insomnia; primary sleep disorders other than primary insomnia; severely disturbed sleep owing to chronic pain, fever, diarrhea, frequent urination, or coughing; organic psychiatric disorders; suicidal ideation or attempt in the past 5 years; clinically severe dysfunction of the liver, kidney, cardiovascular system, or hematologic system or presence of a malignant tumor; and pregnancy or breast-feeding.
All of the subjects completed sleep diaries for 1 week before starting treatment (screening/baseline period). The daily assessment via sleep diary lasted for the first 4 weeks of the treatment period. The information contained in these sleep diaries was used to determine the sleep-related variables used in the analysis (i.e., SL, TST, and WASO). The information contained in the sleep diaries was also used to guide dose escalation. After this initial treatment period, the efficacy of eszopiclone was assessed using the Pittsburgh Sleep Quality Index (PSQI). These data are reported in our prior report.20
The present post hoc analysis excluded patients from the original study with psychiatric disorders (risk of suicide, manic episode, posttraumatic stress disorder, history of or current alcohol dependence or abuse, history of or current drug dependence or abuse, anorexia nervosa, bulimia nervosa, or antisocial personality disorder, as assessed using the Mini-International Neuropsychiatric Interview Japanese version 5.0.025) to avoid confounding results owing to differing responses between patients with psychiatric disorders and those with non-psychiatric disorders.26 Thus, patients who were administered antidepressant or antipsychotic drugs were excluded from the analysis. In addition, the present study focused only on the patients that fell into 2 defined populations in the original study: elderly (age ≥ 65 years) or nonelderly (age < 65 years) patients without psychiatric disorders and who were randomized to 2 mg eszopiclone per day (n = 82). This dose was selected for the present analysis because it was the only dose used by both elderly and nonelderly patients.20 Subjects who had a SL > 30 min at baseline and had evaluable SL data at Week 4 were included in the SL remitter analysis (n = 58). Of these, 56 patients had primary insomnia and 2 had insomnia associated with a physical disorder, namely diabetes. Subjects who had a WASO > 30 min at baseline and had evaluable WASO data at Week 4 were included in the WASO remitter analysis (n = 51). Of these, 48 patients had primary insomnia and 3 had insomnia associated with a physical disorder (diabetes: 2; perimenopause syndrome: 1). The baseline clinical descriptive variables of the subjects are provided in Table 1.
Baseline clinical descriptive variables for the analyzed subjects.
Baseline clinical descriptive variables for the analyzed subjects.
SL remitters in response to eszopiclone were defined as subjects included in the SL remitter analysis whose SL value at Week 4 was ≤ 30 min; that is, patients in whom the SL value had normalized during the 4 weeks of treatment. WASO remitters in response to eszopiclone were defined as subjects in the WASO remitter analysis whose WASO value at Week 4 was ≤ 30 min; that is, patients in whom the WASO value had normalized during the 4 weeks of treatment. These threshold values were set because the presence of SL and WASO values of ≥ 30 min are common complaints of insomnia patients and may indicate a pathological condition.8 They are also commonly used inclusion criteria in studies of insomnia, while normalization of the SL and WASO values to below these threshold levels is commonly used as a measure of response.27–29
The primary efficacy measures that were assessed in the current post hoc analysis were SL and WASO. Week 4 data were used for this analysis because a study duration of 4 weeks is commonly used when evaluating the short-term efficacy of hypnotics.8 The PSQI was also evaluated at baseline and Week 4. All post hoc analyses in the present study were planned after unblinding the study data. Hypnotics and other drugs used to treat insomnia before the trial were classified as prior medications. For patients with prior medications, a washout period of 10 days was set before the start of the trial, and concomitant medications were prohibited during the study period. Drugs that were prohibited for concomitant use from Day 10 prior to the start of treatment until the last treatment day included sedative hypnotics other than eszopiclone, anxiolytics, and strong inhibitors and inducers of cytochrome P-450 3A4.
PSQI scores (total, C2 and C4 scores) at baseline and Week 4 and the 7-day median SL and WASO values in the patient-reported sleep diaries at baseline and Week 4 were used to confirm the adequacy of the evaluation of SL (from C2 scores) or WASO (from C4 scores) from the values in the patients' diaries.
Patients' clinical descriptive variables included gender (number and percentage), age (summary statistics, number and percentage of patients aged < 65 years, and those of patients aged ≥ 65 years), number of current comorbidities (presented as the number and percentage of patients with 0, 1, 2, 3, or 4 or more comorbidities), body mass index (BMI; summary statistics, number and percentage of patients with BMI ≤ 25 kg/m2 or > 25 kg/m2, based on the diagnostic criterion for obesity established by the Japan Society for the Study of Obesity30), and sleep variables (SL, WASO, TST, and PSQI at baseline).
Statistical analyses were performed using SAS ver. 9.2 and 9.3 (SAS Institute Inc, Cary, NC, USA). Because the analysis did not use the last observation carried forward method, dropouts were omitted.
Analysis of the Correlations between PSQI Scores and Both SL and WASO
Spearman rank correlation coefficients for the relationships between PSQI C2 score and SL and between PSQI C4 score and WASO were calculated both at baseline and Week 4.
Analysis of Outcomes in Terms of SL and WASO
Factors associated with a better outcome in response to eszopiclone were examined separately for SL and WASO remitters using logistic regression analysis with a model in which the above-indicated descriptive variables were treated as independent variables. Age (≥ 65 years and < 65 years), BMI (≤ 25 kg/m2 and > 25 kg/m2), and gender were included in the model as categorical variables. The number of current comorbidities, which may affect the development and progression of insomnia symptoms,7,31,32 and pretreatment SL values were also included in the model as continuous variables for the SL remitter analysis. The demographic variables and baseline WASO values were included in the model for WASO remitter analysis.
For the remitter analyses, univariate logistic regression analysis and multivariate logistic regression analysis were performed by including these variables and applying the backward elimination method for variable selection. For these logistic regression analyses, p values were calculated based on Wald χ2 statistics. A two-tailed significance level of 5% was used in the analyses for defining factors associated with improved SL and WASO values in response to eszopiclone treatment.
For each factor that was identified to be associated with the outcome following eszopiclone treatment, a receiver operating characteristic (ROC) curve was generated to determine an arbitrary cutoff value for the determination of remitters. The sensitivity, specificity, and a likelihood ratio (+, −) were calculated for the cutoff values.
Patient Disposition and Baseline Characteristics
Of the 82 elderly and nonelderly patients without psychiatric disorders who were treated with 2 mg eszopiclone, 78 patients had evaluable data at Week 4 and were included in the analysis. Week 4 data were unavailable for the remaining 4 patients owing to premature discontinuation or missing data; therefore, they were excluded from the analysis. SL remitter analysis was performed on data from 58 patients in whom the median baseline SL was > 30 min over a period of 1 week before treatment started and who had evaluable SL data at Week 4. In the other 20 patients, the median SL over 1 week was ≤ 30 min despite the inclusion criterion of SL ≥ 30 min on ≥ 3 of 7 days. WASO remitter analysis was performed on data from 51 patients in whom the baseline WASO was > 30 min and who had evaluable WASO data at Week 4. Twenty-seven patients with a 1-week median WASO ≤ 30 min were excluded. Patient disposition in the original study for the post hoc analyses performed in the present study are shown in Figure 1.
The study included elderly and nonelderly patients without psychiatric disorders who were treated with 2 mg eszopiclone per day (n = 82). Of these, 78 had evaluable data at Week 4 and were included in the analysis. The SL remitter analysis was performed in 58 patients (28 elderly and 30 nonelderly) who had evaluable SL data at Week 4 and a median baseline SL > 30 min. The WASO remitter analysis was performed in 51 patients (25 elderly and 26 nonelderly) who had evaluable WASO data at Week 4 and a median baseline WASO > 30 min.
Patient disposition.The study included elderly and nonelderly patients without psychiatric disorders who were treated with 2 mg eszopiclone per day (n = 82). Of these, 78 had evaluable data at Week 4 and were included in the analysis. The SL remitter analysis was performed in 58 patients (28 elderly and...
The mean baseline PSQI (SD) of subjects (n = 77) was 11.5 (2.2), indicating moderate-to-severe disease severity in our subjects (Table 1). Three elderly patients were diagnosed with insomnia associated with physical disorders, which consisted of diabetes in 2 and menopausal syndrome in the other. The remaining patients had primary insomnia. The correlations (Spearman rank correlation coefficient) at baseline between PSQI C2 and SL, and between PSQI C4 and WASO, were 0.501 (p < 0.001) and 0.599 (p < 0.001) (n = 78), respectively.
Associations of Clinical Descriptive Variables with SL or WASO Improvements
Of 58 subjects included in the SL remitter analysis, 35 (60.3%) were remitters. The SL remitter rate was considerably higher among females (23 of 33 female subjects [69.7%] were SL remitters) than among males (12 of 25 males [48.0%] were SL remitters), but the difference was not statistically significant in either univariate (p = 0.098) or multivariate (p = 0.175) analyses, as described below. The SL remitter rate was similar between subjects aged < 65 years (16 of 30 subjects [53.3%]) and subjects aged ≥ 65 years (19 of 28 subjects [67.9%]). Of 51 subjects included in the WASO remitter analysis, 35 (68.6%) were remitters. Similar to the changes in SL, the WASO remitter rate was higher among females (22 of 30 female subjects [73.3%] were WASO remitters) than among males (13 of 21 males [61.9%] were WASO remitters), but again the difference was not statistically significant (univariate analysis, p = 0.389; multivariate analysis, p = 0.305). The WASO remitter rate was similar between subjects aged < 65 years (18 of 26 subjects [69.2%]) and subjects aged ≥ 65 years (17 of 25 subjects [68.0%]). There was no difference in age between remitters and non-remitters for SL or WASO.
No significant associations were found between SL or WASO improvements and gender, age, BMI, or the number of current comorbidities in either the univariate or multivariate logistic regression analyses (Tables 2 and 3). However, baseline SL showed a significant association with SL improvement in both univariate (p = 0.026) and multivariate (p = 0.037) analyses. Likewise, baseline WASO was significantly associated with WASO improvement in the univariate (p = 0.001) and multivariate (p < 0.001) analyses. In support of these findings, in the variable selection models, only baseline SL was included in the final model for SL remitters while only baseline WASO was included in the final model for WASO remitters.
Logistic regression analyses for factors associated with SL remitters.
Logistic regression analyses for factors associated with SL remitters.
Logistic regression analyses for factors associated with WASO remitters.
Logistic regression analyses for factors associated with WASO remitters.
Receiver Operating Characteristic Curves
ROC curves for SL remitters and WASO remitters are shown in Figure 2. SL of 75 min at baseline was selected as the arbitrary cutoff value for determination of SL remitters/non-remitters. The sensitivity and specificity for this cutoff value were 0.800 and 0.609, respectively. WASO of 80 min at baseline was selected as the arbitrary cutoff value for determination of WASO remitters/non-remitters. The sensitivity and specificity for this cutoff value were 0.800 and 0.938, respectively.
Receiver operating characteristic curves.
(A) Sleep latency remitters and (B) wake after sleep onset remitters.
Receiver operating characteristic curves. (A) Sleep latency remitters and (B) wake after sleep onset remitters.
In the present study, the mean PSQI total score in the subjects was 11.5, which was previously classified as moderate insomnia severity.33 Considering that only a small number of patients (three elderly patients) were diagnosed with insomnia associated with physical disorders, the result of the present study mostly represents a treatment response to eszopiclone in patients with primary insomnia. Although a dose-finding study in patients with primary insomnia conducted in the United States showed greater reductions in WASO with 3 mg eszopiclone compared with 2 mg eszopiclone in nonelderly patients,12 only patients who received 2 mg eszopiclone in the original trial were included in the present post hoc analysis. This enabled us to investigate the differences in treatment outcomes in response to a common dose of eszopiclone between elderly and nonelderly patients. The medium to strong correlations between PSQI sub-item scores for evaluating sleep onset disturbance and SL, and between PSQI sub-item scores for evaluating sleep maintenance disturbance and WASO at baseline, demonstrate the appropriateness of using the baseline SL and WASO values (both of which were obtained from the patients' sleep diaries) as evaluation variables.
The proportions of SL and WASO remitters in the present study were approximately 60% and 70%, respectively, suggesting that a considerable number of patients do not respond sufficiently to treatment with 2 mg eszopiclone. However, no relationships between therapeutic outcomes and demographic factors were identified. Previous reports have described relationships between patient demographic factors and increased likelihood of insomnia, including a high incidence of insomnia in elderly people and women.21,22,34,35 However, the results of the present study showed no age- or gender-dependent difference in the outcomes in response to eszopiclone, which suggests that the effectiveness of eszopiclone is not different between genders or between elderly/non-elderly patients. It is particularly interesting that age did not influence the outcomes in response to eszopiclone even though it was reported to affect the pharmacokinetic properties of eszopiclone.14 Interestingly, there was no relationship between the number of current comorbidities, which has been reported to be related to the development and worsening of insomnia,7,31,32 and treatment outcomes in the present analysis. Thus, eszopiclone could improve insomnia symptoms regardless of the patients' background conditions.
On the other hand, the present results do reveal associations between treatment outcomes (based on SL and WASO only) and baseline sleep parameters. Indeed, both WASO and SL remitters could be identified using WASO and SL values at baseline, respectively. For sleep-onset insomnia, the therapeutic response to eszopiclone was associated with baseline SL, and a normalization of sleep-onset insomnia was verified in patients with SL ≤ 75 min, which is within the range of moderate severity.17 Similarly, subjective WASO at baseline was related to the treatment outcome. In addition, the ROC analysis revealed that a WASO of 80 min was the cutoff level for WASO remitters. The relationship between the duration of WASO and the subjective severity of insomnia has not been evaluated in detail. However, in the baseline data of the placebo group in a 6-month administration study of eszopiclone in adult patients with primary insomnia,17 WASO was around 40 min and the insomnia severity index (ISI) score was 17.8, which indicates moderate severity. A severity classification based on the ISI score revealed that patients with moderate insomnia (defined as an ISI score of 15–21) accounted for 62.9% of the patients, while those with severe insomnia (ISI scores of 22–28) accounted for 17.5% of the patients.17 Therefore, it seems reasonable to classify patients with a baseline WASO > 80 min, which was estimated to be a risk factor for WASO non-response in the present analysis, as relatively severe insomnia cases.
The results of the present study suggest that 2 mg eszopiclone per day is effective for patients with mild-to-moderate sleep-onset insomnia or sleep maintenance insomnia. In the present study, the relatively short SL or WASO values were associated with the favorable outcome of eszopiclone treatment. These findings differ from those of previous studies examining the predictors of the response to cognitive behavioral therapy.36,37 This phenomenon might be due to the different mechanisms of action between these two treatments modalities because eszopiclone displays hypno-sedative properties whereas the main component of cognitive behavioral techniques is sleep compression, which is particularly effective in terms of reducing sleep latency.37
The findings of the present analysis should be considered in light of the study's limitations. First, it is not clear whether the patients in this study are representative of patients with insomnia in general clinical practice because they were sampled from a clinical trial. However, because the PSQI scores in our patients were within the range of those in the population with chronic insomnia reported by Buysse et al.,38 we believe that they could be considered representative. Second, because there were no clearly specified criteria for early-morning awakening and interruptions of nocturnal sleep in the trial, these symptoms were not evaluated in the present study. The outcomes in response to treatment of these symptoms should be evaluated in future studies, as should the difference in cutoff values between remitters in response to 3 or 2 mg eszopiclone. Third, because information about the duration of illness could not be obtained, it was not possible to specify whether the insomnia was chronic or acute. Although most of the patients were considered to have chronic insomnia (because they met the DSMIV diagnostic criterion for primary insomnia that progresses chronically, and they had experienced persistent insomnia for ≥ 4 weeks at the time of inclusion), the possibility that the therapeutic responses may vary depending on the duration of illness should be examined. Fourth, the thresholds for WASO and SL used (> 30 min) for inclusion in the study are not definitive, but rather just common complaints of insomnia patients, and so may not be the most appropriate values for a post hoc analysis. Fifth, in general, hypnotics are known to have a considerable placebo effect, accounting for up to half of the drug response.39 Because the original trial used for this study was not placebo controlled, it was not possible to assess the placebo effect; that is, to determine what proportion of the drug response was simply an improvement that might have occurred naturally. Sixth, the sleep measures were subjectively assessed, being based on information recorded in patients' sleep diaries; objective measures such as polysomnography and actigraphy were not used. Therefore, we hope to verify the present results in a future study using polysomnography. Seventh, we included patients who reported that their primary insomnia symptoms consisted of difficulty in maintaining sleep or difficulty initiating sleep, and these symptoms may coexist or change over time. In this study, we used subjective assessments of SL and WASO to predict the clinical outcomes in these patients, but these may not be sufficiently sensitive in patients with both of these symptoms or whose symptoms change over time. Finally, the exclusion of patients with psychiatric disorders—specifically anxiety and depression, which are commonly comorbid with insomnia—from the present analysis means that our sample may not be representative of patients with insomnia in general clinic practice. However, we excluded patients with psychiatric disorders to avoid confounding results due to differing responses between patients with psychiatric disorders and those with non-psychiatric disorders.26
In summary, in the present study, the proportions of SL and WASO remitters were in the range of 60% to 70% during treatment with 2 mg eszopiclone per day. SL at baseline was related to SL improvement, while WASO at baseline was related to WASO improvement. Eszopiclone was sufficiently effective for treating both sleep-onset insomnia and waking after sleep onset in patients with symptoms of mild to moderate severity. These findings may be helpful for guiding decisions about the appropriate dose of hypnotics, and may help clinicians to predict their patients' outcomes in response to hypnotics prior to the start of treatment. However, the sample number was too small to perform further stratification of patients. Because this analysis was conducted in a post hoc retrospective manner, prospective studies involving greater numbers of patients are needed to verify our results.
Dr. Yuichi Inoue received financial support for developing the manuscript, and personal fees for providing expert testimony and lectures including service on speakers bureaus from Eisai Co., Ltd.; personal fees for providing expert testimony and lectures including service on speakers bureaus from Otsuka Pharmaceutical Co., Ltd.; personal fees for providing expert testimony and lectures including service on speakers bureaus from Pacific Medico Co., Ltd. and Mitsubishi Tanabe Pharma Corporation; and personal fees for lectures including service on speakers bureaus from GlaxoSmithKline K.K., Astellas Pharma Inc., Sanofi-Aventis K.K., Yoshitomiyakuhin Corporation, Nippon Boehringer Ingelheim Co., Philips Respironics G.K., Alfresa Pharma Corporation, Takeda Pharmaceutical Company Ltd., and MSD K.K. Dr. Inoue also has a financial interest/relationship with Hisamitsu Pharmaceutical Co. regarding an investigational device/drug. Dr. Inoue's research team has also received Health and Labour Sciences Research and Comprehensive Research on Disability Health and Welfare grants. Atsushi Kamijo and Reiko Nagai are employees of Eisai Co., Ltd.
body mass index
Diagnostic and Statistical Manual of Mental Disorders
insomnia severity index
Pittsburgh Sleep Quality Index
receiver operating characteristic
total sleep time
wake time after sleep onset
The authors thank Daniel McGowan, PhD, and Nicholas Smith, PhD, for providing medical writing support in the preparation of this manuscript.
Morin CM, Bélanger L, LeBlanc M, et al., authors. The natural history of insomnia: a population-based 3-year longitudinal study. Arch Intern Med. 2009;169:447–3. [PubMed]
American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 5th ed. Washington: American Psychiatric Association, 2013.
Ohayon MM, author. Epidemiology of insomnia: what we know and what we still need to learn. Sleep Med Rev. 2002;6:97–111. [PubMed]
Yokoyama E, Kaneita Y, Saito Y, et al., authors. Association between depression and insomnia subtypes: a longitudinal study on the elderly in Japan. Sleep. 2010;33:1693–702. [PubMed Central][PubMed]
Sasai T, Inoue Y, Komada Y, Nomura T, Matsuura M, Matsushima E, authors. Effects of insomnia and sleep medication on health-related quality of life. Sleep Med. 2010;11:452–7. [PubMed]
Foley DJ, Monjan A, Simonsick EM, Wallace RB, Blazer DG, authors. Incidence and remission of insomnia among elderly adults: an epidemiologic study of 6,800 persons over three years. Sleep. 1999;22:S366–72. [PubMed]
Foley D, Ancoli-Israel S, Britz P, Walsh J, authors. Sleep disturbances and chronic disease in older adults: results of the 2003 National Sleep Foundation Sleep in America Survey. J Psychosom Res. 2004;56:497–502. [PubMed]
Schutte-Rodin S, Broch L, Buysse D, Dorsey C, Sateia M, authors. Clinical guideline for the evaluation and management of chronic insomnia in adults. J Clin Sleep Med. 2008;4:487–504. [PubMed Central][PubMed]
Zammit GK, McNabb LJ, Caron J, Amato DA, Roth T, authors. Efficacy and safety of eszopiclone across 6-weeks of treatment for primary insomnia. Current Med Res Opin. 2004;20:1979–91.
Scharf M, Erman M, Rosenberg R, et al., authors. A 2-week efficacy and safety study of eszopiclone in elderly patients with primary insomnia. Sleep. 2005;28:720–7. [PubMed]
McCall WV, Erman M, Krystal AD, et al., authors. A polysomnography study of eszopiclone in elderly patients with insomnia. Current Med Res Opin. 2006;22:1633–42.
Erman MK, Zammit G, Rubens R, et al., authors. A polysomnographic placebo-controlled evaluation of the efficacy and safety of eszopiclone relative to placebo and zolpidem in the treatment of primary insomnia. J Clin Sleep Med. 2008;4:229–34. [PubMed Central][PubMed]
Lam SP, Fong SY, Ho CK, Yu MW, Wing YK, authors. Parasomnia among psychiatric outpatients: a clinical, epidemiologic, cross-sectional study. J Clin Psychiatry. 2008;69:1374–82. [PubMed]
McCrae CS, Ross A, Stripling A, Dautovich ND, authors. Eszopiclone for late-life insomnia. Clin Interv Aging. 2007;2:313–26. [PubMed Central][PubMed]
Roth T, Walsh JK, Krystal A, Wessel T, Roehrs TA, authors. An evaluation of the efficacy and safety of eszopiclone over 12 months in patients with chronic primary insomnia. Sleep Med. 2005;6:487–95. [PubMed]
Soares CN, Joffe H, Rubens R, Caron J, Roth T, Cohen L, authors. Eszopiclone in patients with insomnia during perimenopause and early postmenopause: a randomized controlled trial. Obstet Gynecol. 2006;108:1402–10. [PubMed]
Walsh JK, Krystal AD, Amato DA, et al., authors. Nightly treatment of primary insomnia with eszopiclone for six months: effect on sleep, quality of life, and work limitations. Sleep. 2007;30:959–68. [PubMed Central][PubMed]
Roth T, Price JM, Amato DA, Rubens RP, Roach JM, Schnitzer TJ, authors. The effect of eszopiclone in patients with insomnia and coexisting rheumatoid arthritis: a pilot study. Prim Care Companion J Clin Psychiatry. 2009;211:292–301.
Greenblatt DJ, Zammit GK, authors. Pharmacokinetic evaluation of eszopiclone: clinical and therapeutic implications. Expert Opin Drug Metab Toxicol. 2012;8:1609–18. [PubMed]
Uchimura N, Kamijo A, Takase T, authors. Effects of eszopiclone on safety, subjective measures of efficacy, and quality of life in elderly and nonelderly Japanese patients with chronic insomnia, both with and without comorbid psychiatric disorders: a 24-week, randomized, double-blind study. Ann Gen Psychiatry. 2012;11:15. [PubMed Central][PubMed]
López-Torres Hidalgo J, Navarro Bravo B, Párraga Martínez I, Andrés Pretel F, Téllez Lapeira J, Boix Gras C, authors. Understanding insomnia in older adults. Int J Geriatr Psychiatry. 2012;27:1086–93. [PubMed]
Zhang B, Wing YK, authors. Sex differences in insomnia: a meta-analysis. Sleep. 2006;29:85–93. [PubMed]
American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th ed. Washington: American Psychiatric Association, 1994.
Fava M, McCall WV, Krystal A, et al., authors. Eszopiclone co-administered with fluoxetine in patients with insomnia coexisting with major depressive disorder. Biol Psychiatry. 2006;59:1052–60. [PubMed]
Otsubo T, Miyaoka H, Kamijima K, authors. Mini-International Neuropsychiatric Interview (M.I.N.I.) Japanese version 5.0.0. 2003. Tokyo: Seiwashoten. [In Japanese].
Dombrovski AY, Mulsant BH, Houch PR, et al., authors. Residual symptoms and recurrence during maintenance treatment of late-life depression. J Affect Disord. 2007;103:77–82. [PubMed Central][PubMed]
Watanabe N, Furukawa TA, Shimodera S, et al., authors. Brief behavioral therapy for refractory insomnia in residual depression: an assessor-blind, randomized controlled trial. J Clin Psychiatry. 2011;72:1651–8. [PubMed]
Belleville G, Guay C, Guay B, Morin CM, authors. Hypnotic taper with or without self-help treatment of insomnia: a randomized clinical trial. J Consult Clin Psychol. 2007;75:325–35. [PubMed]
Epstein DR, Sidani S, Bootzin RR, Belyea MJ, authors. Dismantling multicomponent behavioral treatment for insomnia in older adults: a randomized controlled trial. Sleep. 2012;35:797–805. [PubMed Central][PubMed]
Takahashi H, Mori M, authors. Characteristics and significance of criteria for obesity disease in Japan 2011. Nihon Rinsho. 2013;71:257–61. [In Japanese]. [PubMed]
Foley DJ, Vitiello MV, Bliwise DL, Ancoli-Israel S, Monjan AA, Walsh JK, authors. Frequent napping is associated with excessive daytime sleepiness, depression, pain, and nocturia in older adults: findings from the National Sleep Foundation ‘2003 Sleep in America’ Poll. Am J Geriatr Psychiatry. 2007;15:344–50. [PubMed]
Sarsour K, Morin CM, Foley K, Kalsekar A, Walsh JK, authors. Association of insomnia severity and comorbid medical and psychiatric disorders in a health plan-based sample: insomnia severity and comorbidities. Sleep Med. 2010;11:69–74. [PubMed]
Komada Y, Nomura T, Kusumi M, et al., authors. A two-year follow-up study on the symptoms of sleep disturbances/insomnia and their effects on daytime functioning. Sleep Med. 2012;13:1115–21. [PubMed]
Kim K, Uchiyama M, Okawa M, Liu X, Ogihara R, authors. An epidemiological study of insomnia among the Japanese general population. Sleep. 2000;23:41–7. [PubMed]
Krystal AD, author. Insomnia in women. Clin Cornerstone. 2003;5:41–50.
Espie CA, Inglis SJ, Harvey L, authors. Predicting clinically significant response to cognitive behavior therapy for chronic insomnia in general medical practice: analysis of outcome data at 12 months posttreatment. J Consult Clin Psychol. 2001;69:58–66. [PubMed]
Troxel WM, Conrad TS, Germain A, Buysse DJ, authors. Predictors of treatment response to brief behavioral treatment of insomnia (BBTI) in older adults. J Clin Sleep Med. 2013;9:1281–9. [PubMed Central][PubMed]
Buysse DJ, Reynolds CF 3rd, Monk TH, Berman SR, Kupfer DJ, authors. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res. 1989;28:193–213. [PubMed]
Huedo-Medina TB, Kirsch I, Middlemass J, Klonizakis M, Siriwardena AN, authors. Effectiveness of non-benzodiazepine hypnotics in treatment of adult insomnia: meta-analysis of data submitted to the Food and Drug Administration. BMJ. 2012;345:e8343. [PubMed Central][PubMed]