Randomized Study on the Efficacy of Cognitive-Behavioral Therapy for Insomnia Secondary to Breast Cancer, Part I: Sleep and Psychological Ef
http://www.100md.com
《临床肿瘤学》
the Laval University Cancer Research Center
School of Psychology, Université Laval, Québec, Québec, Canada
ABSTRACT
PURPOSE: Chronic insomnia is highly prevalent in cancer patients. Cognitive-behavioral therapy (CBT) is considered the treatment of choice for chronic primary insomnia. However, no randomized controlled study has been conducted on its efficacy for insomnia secondary to cancer. Using a randomized controlled design, this study conducted among breast cancer survivors evaluated the effect of CBT on sleep, assessed both subjectively and objectively, and on hypnotic medication use, psychological distress, and quality of life.
PATIENTS AND METHODS: Fifty-seven women with insomnia caused or aggravated by breast cancer were randomly assigned to CBT (n = 27) or a waiting-list control condition (n = 30). The treatment consisted of eight weekly sessions administered in a group and combined the use of stimulus control, sleep restriction, cognitive therapy, sleep hygiene, and fatigue management. Follow-up evaluations were carried out 3, 6, and 12 months after the treatment.
RESULTS: Participants who received the insomnia treatment had significantly better subjective sleep indices (daily sleep diary, Insomnia Severity Index), a lower frequency of medicated nights, lower levels of depression and anxiety, and greater global quality of life at post-treatment compared with participants of the control group after their waiting period. Results were more equivocal on polysomnographic indices. Therapeutic effects were well maintained up to 12 months after the intervention and generally were clinically significant.
CONCLUSION: This study supports the efficacy of CBT for insomnia secondary to breast cancer.
INTRODUCTION
Insomnia occurs frequently in the context of cancer, affecting 30% to 50% of cancer patients.1 Insomnia is particularly frequent in breast cancer patients relative to other cancer types.2 A recent survey3 revealed that 51% of women treated for breast cancer reported nonspecific sleep difficulties and 19% met diagnostic criteria for an insomnia syndrome, which was chronic in 95% of the patients (insomnia duration of 6 months or more). Insomnia is associated with significant burden, both for the individual and for the health care system, in terms of functional impairments, reduced quality of life, and increased health care use,4 hence the importance of treating insomnia, particularly when it is chronic.
Psychological interventions should be the preferred alternative, compared with hypnotic medications, when insomnia is chronic.5-8 The efficacy of psychological interventions, especially cognitive-behavioral therapy (CBT), is well established for primary insomnia (ie, insomnia not etiologically related to a medical or a psychological condition). There is also recent evidence supporting the efficacy of CBT for insomnia secondary to chronic pain9 and to a medical or a psychological condition,10 as well as for insomnia comorbid with diverse medical conditions.11 However, studies conducted on insomnia associated with cancer are scarce.
Early pilot studies yielded mixed results on the efficacy of relaxation for insomnia in cancer patients.12,13 More recently, nonrandomized studies showed positive outcomes associated with multimodal psychological interventions for insomnia, including improved subjective sleep and some aspects of quality of life.14,15 We also found promising results in a pilot study using a multiple-baseline design, a more stringent definition of secondary insomnia, and objective sleep measures.16 This study conducted among eight women with insomnia caused or aggravated by breast cancer showed reduced total wake time and increased sleep efficiency, both assessed subjectively (ie, sleep diary) and objectively (ie, polysomnography), as well as improved mood, fatigue, and quality of life. However, a randomized trial is still needed to document more systematically the efficacy of CBT for insomnia secondary to cancer.
The goals of this study were to assess the efficacy of CBT for insomnia secondary to breast cancer, in a randomized controlled group design, on measures of subjective and objective sleep, psychological functioning (ie, depression, anxiety, fatigue), and quality of life. The study also investigated the effect of CBT for insomnia on immunologic functioning. These results will be reported in a companion article because of space limitations (Savard et al17).
PATIENTS AND METHODS
Participants
The participants were recruited between November 1998 and December 2001 with fliers and pamphlets, ads placed in the local newspapers, and by physician referrals. To be accepted in the study, participants had to have completed radiotherapy and chemotherapy for a stage I to III breast cancer at least 1 month prior to enrollment onto the study, and had to meet diagnostic criteria for a chronic insomnia syndrome, as defined by the combined criteria of the International Classification of Sleep Disorders18 and of the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition (DSM-IV),19 as well as those commonly used in insomnia research.20 Those criteria included difficulty in initiating and/or maintaining sleep, whereby sleep-onset latency and/or wake after sleep onset is greater than 30 minutes; sleep efficiency (ratio of total sleep time to total time spent in bed) lower than 85%; difficulties occurring at least 3 nights per week; difficulties occurring for at least 6 months; and difficulties causing marked distress or significant impairment in daytime functioning (eg, fatigue, disturbed mood, performance deficits). Only patients whose insomnia was judged to be secondary to cancer were included in the study (ie, those whose sleep difficulties were caused or aggravated by the cancer diagnosis or treatment).
The following exclusion criteria were also used: presence of severe major depression or another serious psychiatric disorder; presence of a sleep disorder other than insomnia (eg, sleep apnea, periodic limb movements); presence of another illness affecting the immune system (eg, HIV infection); regular use of a psychotropic medication other than hypnotics (eg, antidepressants), unless the dosage used was stable in the last month and did not increase during the study; and current involvement in psychotherapy.
Patients using a stable dosage of hypnotic medication were not excluded from the study, provided that they met all criteria described, including the diagnostic criteria for an insomnia syndrome. Thus, hypnotic users who were included in the study were all reporting clinically significant insomnia despite the use of this medication (hence suggesting its limited utility). Including those patients has the advantage of increasing the study's external validity. Indeed, it has been found that 20% to 40% of cancer patients use a psychotropic medication for sleep difficulties.2,21,22 This variable (changes in hypnotic use) was used as an outcome measure in this study. Several other studies conducted in the context of primary23,24 or secondary9,11 insomnia have included patients using a hypnotic medication and used this variable as an outcome measure. Statistical analyses were nonetheless performed to ensure the inclusion of these patients did not influence the effect of the intervention on other sleep variables (see Statistical Analyses).
Sample characteristics. A priori power analyses, based on effect sizes obtained from two meta-analyses assessing the efficacy of psychological interventions for primary insomnia,25,26 yielded a total sample size estimate for this study ranging from 50 to 60 participants. Figure 1 shows rates of recruitment, exclusion, refusal, and dropout throughout the study. As illustrated, of the 160 women who responded to the recruitment effort, 58 (36.3% of women who were initially screened over the phone) were enrolled in the study and randomly assigned to the treatment or the control condition. Demographic and clinical characteristics of the final sample of 57 participants at study entry are listed in Table 1. All patients were white. Of all demographic and clinical variables examined, the only group difference at pretreatment was for the proportion of patients with a comorbid physical illness (eg, cardiovascular disease, arthritis), which was greater in the treatment (59%) than in the control condition (27%; 21 = 6.19; P < .05).
Experimental Design
Participants were randomly assigned to one of the two following conditions: CBT administered immediately or waiting-list control condition (WLC; Fig 2). Because the recruitment was somewhat slow and treatment was conducted in small groups, the random allocation of patients was done by blocks of five to six patients. Participants assigned to the control group waited a minimum of 8 weeks, which corresponded to the duration of the intervention, were assessed again on study variables, including completion of a sleep diary for a 2-week period, and then received CBT. This second assessment (pretreatment or postwaiting assessment) of control patients was contrasted with the post-treatment evaluation of treated patients to verify the short-term effect of CBT. Additional evaluations were conducted 3, 6, and 12 months after the end of their respective treatment to assess the maintenance of treatment effects over time. The study was approved by the ethical review boards of L'H?tel-Dieu de Québec and Université Laval (Québec, Canada).
Measures
Insomnia Interview Schedule. The Insomnia Interview Schedule (IIS)20 consists of a series of semistructured questions evaluating several aspects of sleep difficulties. It also provides guidelines for diagnosing insomnia according to the criteria described and for detecting symptoms suggestive of other sleep disorders (eg, obstructive sleep apnea, periodic limb movement disorder). For the purpose of this study, questions were added to the IIS to determine whether insomnia was secondary to breast cancer by assessing the temporal relationship between the cancer diagnosis (or treatment) and the onset or aggravation of sleep disturbances.10,27,28
Structured Clinical Interview for the DSM-IV. The Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (SCID),29 is used to evaluate the presence of current and past psychiatric disorders according to DSM-IV diagnostic criteria.19
Sleep diary. The sleep diary20 (subjective sleep measure) was completed each day on arising and provided information pertaining to the use of hypnotic medications as sleep aids, bedtime hour, time of falling asleep, number and duration of awakenings during the night, and the time of morning awakening and arising from bed. The dependent variables derived from that information were sleep onset latency (time to sleep after lights out), wake after sleep onset (summation of nocturnal awakenings), total wake time (summation of sleep onset latency, wake after sleep onset, and early morning awakening), total sleep time (time in bed minus total wake time), sleep efficiency (ratio of total sleep time to the actual time spent in bed multiplied by 100), and use of sleep-promoting medications (type, frequency, and quantity). Although subjective reports do not always correspond to polysomnographic sleep data, they still provide a reliable index of insomnia.30
Polysomnography. The polysomnography (PSG; objective sleep measure) montage included standard electroencephalographic, electromyographic, and electro-oculographic monitoring. On the first night in the laboratory, respiration (airflow, tidal volume, and oxygen saturation) and anterior tibialis electromyography were also measured to rule out the presence of sleep apnea or periodic limb movements. The sleep stages were scored in 30-second epochs, using standard criteria31 by an experienced technician who was blinded to the patient's experimental condition. The following dependent variables were derived from the polysomnographic data: sleep onset latency (time from lights out to persistent sleep), wake after sleep onset (total of nocturnal awakenings), total wake time (latency to sleep + wake after sleep onset + early morning awakening), total sleep time (time in bed minus total wake time), and sleep efficiency (ratio of total sleep time to the actual time spent in bed multiplied by 100). These sleep variables were averaged from nights 2 and 3 for each study time assessment, with night 1 excluded to allow for patients' adaptation to the laboratory setting.
Insomnia Severity Index. The Insomnia Severity Index (ISI)20 is a seven-item questionnaire designed to evaluate insomnia severity on the basis of difficulties falling asleep, night-time awakenings, early morning awakenings, impairment of daytime functioning due to sleep problems, noticeability of impairments, distress or worry caused by sleep difficulties, and dissatisfaction with sleep. Each item is rated using a five-point Likert scale ranging from 0 (not at all) to 4 (very much), for a total score ranging from 0 to 28. In this study, three parallel versions of the ISI were administered: one that was completed by the participant (ISI-P), one that was completed by the clinician (ISI-C), and one that was completed by a participant's significant other (ISI-SO). We recently validated the French-Canadian version of the ISI specifically in the context of cancer,32 and found it had psychometric properties equivalent to those in the original version.33,34
Hospital Anxiety and Depression Scale. This questionnaire includes 14 items divided into two subscales: depression (Hospital Anxiety and Depression Scale [HADS] -D35: seven items) and anxiety (HADS-A: seven items). A score of 7 and more, for a maximum score of 21, at each subscale suggests the presence of clinical levels of depression or anxiety.36,37 The French-Canadian version proved to possess psychometric qualities equivalent to those of the original English version.38
Multidimensional Fatigue Inventory. The French-Canadian version of the Multidimensional Fatigue Inventory (MFI),39 which possesses adequate psychometric properties,40 is a short form of the original MFI. It contains 15 items (on a scale from 1 to 5) divided into four subscales including general and physical fatigue, reduction in activities, reduction in motivation, and mental fatigue. A global score of fatigue is obtained by calculating a mean for all items.
The European Organization for Research and Treatment of Cancer Quality of Life Questionnaire. The European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (QLQ-C30+3)41,42 was developed and validated with cancer patients. Only results obtained on the global quality-of-life scale, comprised of three items, are reported in this article. Scores are transformed to give a score ranging from 0 to 100. The French version was developed by the authors of the original English version.
Procedure
Telephone screening. A telephone screening was undertaken with women seeking information about the study to evaluate their eligibility briefly and explain the study goals and procedures. A clinical interview was scheduled with interested and eligible patients based on initial criteria (eg, sleep difficulties at least 3 nights a week for at least 6 months).
Clinical interview. During this interview, study procedures were explained in detail to participants and their written consent was obtained. Then, the IIS and the SCID were administered individually by two trained and independent clinicians. The clinician who administered the IIS also completed the ISI-C after the interview. After the evaluations, the two clinicians met to discuss the information obtained and to decide on the eligibility of the participant. Eligible participants were asked to complete a daily sleep diary for a period of 2 weeks. At this point, patients were considered eligible for the study when information of sleep diaries confirmed data obtained with the IIS concerning the frequency (ie, at least 3 nights per week) and severity (ie, awakenings > 30 minutes, sleep efficiency < 85%) of sleep difficulties.
PSG. After the clinical interview, participants spent 3 consecutive nights in a sleep laboratory for PSG assessment. The day after, patients with no evidence of sleep apnea or periodic limb movement based on PSG were definitely accepted onto the study (which was the case for all patients) and randomly assigned to treatment or control conditions. Patients allocated to the control group completed the battery of self-report scales (ISI-P, HADS, MFI, and QLQ-C30+3) and were instructed to ask a significant other to complete the ISI-SO, which provided the prewaiting measures.
Information session. Before beginning the intervention, participants were invited to an information session explaining the general treatment procedures. All participants also had to complete the same battery of self-report scales, which provided the pretreatment measures, and to ask a significant other to complete the ISI-SO.
Treatment. The insomnia treatment consisted of eight weekly sessions of approximately 90 minutes, offered in groups of four to six patients. The treatment protocol was based on clinical procedures developed by Morin20 and slightly adapted by our team for the cancer population. This multimodal approach combined behavioral (ie, stimulus control therapy, sleep restriction), cognitive (ie, cognitive restructuring), and educational (ie, sleep hygiene, fatigue and stress management) strategies that were described in a treatment manual given to all participants. Participants were instructed to continue completing the daily sleep diary throughout the treatment for clinical purposes. The treatment content is described in more detail elsewhere.16
The treatment was administered by a master-level psychologist with experience in the administration of this particular treatment protocol. An optional booster session was offered to participants 1 month after the end of the treatment. Participants interested in reducing their use of sleep-promoting medication were advised to consult a physician or pharmacist for supervision of medication withdrawal. Because missed treatment sessions (n = 8) were rescheduled, all patients received the entire treatment program, excluding the four patients who dropped out of the study during the course of the intervention (Fig 1).
Post-treatment procedure and follow-up. After the intervention, as well as 3, 6, and 12 months after the end of treatment, the participants again completed the battery of self-report scales and completed daily sleep diaries for another 2-week period. The clinician and the significant other also rated their respective versions of the ISI at these times. In addition, at post-treatment and 6-month follow-up, the participants spent 3 consecutive nights in the laboratory to measure therapeutic progress objectively.
Statistical Analyses
All data were carefully inspected to identify missing data and outliers, and to assess normality.43 Descriptive and inferential statistics were completed using SAS 8.2 statistical software (SAS Institute, Cary, NC).44 For all inferential tests, = .05 (two tailed). The main analyses were based on a split-plot group (two conditions) -time (five assessments: pretreatment, post-treatment, and 3-, 6-, and 12-month follow-up) randomized design. Two subsets of analyses were performed. First, analyses were conducted to determine whether treated patients had greater improvements on all dependent variables at post-treatment compared with patients in the control group after their waiting period. These findings are reported in Group Comparisons; because of this objective, only significant group-time interactions are reported. Then, because of the nonsignificant group effects obtained, data of both groups were pooled together to confirm, with a larger sample size, the benefits associated with the intervention at post-treatment, and to evaluate whether therapeutic gains observed at post-treatment were maintained at follow-up assessments. These findings are reported in Pooled Analyses and significant time effects are emphasized. A secondary analysis was also performed to investigate the moderating role of hypnotic use at pretreatment on the effect of CBT on subjective sleep measures (sleep diary and ISI). This was done by testing whether the third-order interaction (hypnotic use-group-time) was significant.45
Data were analyzed within an intent-to-treat framework. Linear mixed models were used to test group, time, and group-time interaction effects for all continuous dependent variables. A priori contrasts were used to break down these effects. Satterthwaite F tests were computed because they are typically more robust to non-normality, unbalanced data, and violations of multisample sphericity.46
In accordance with the strategy suggested by Frigon and Laurencelle,47 various covariates (eg, hormone therapy and the presence of a comorbid medical condition) were tested to assess their capacity to reduce the error term. To be included in the mixed-model analyses as a covariate, a variable had to meet these two criteria: significant between-group differences and significant reduction of error variance of more than one dependent variable. As mentioned, significant between-group differences were found on only one variable—the presence of a comorbid illness—but this variable was only significantly associated with total sleep time as assessed with the sleep diary. Hence, results reported thereafter are those based on analyses conducted without covariates. For analyses conducted on PSG data, however, five covariates were added to the mixed models to control for the potential impact on sleep of various health behaviors the patients had during the 24 hours preceding each recording night (ie, caffeine and alcohol consumption, sleep medication use, napping, and pain level).
Finally, generalized estimating equations (SAS CATMOD procedure; SAS Institute) analyses were used to estimate the three main effects (group, time, and group-time interaction) for the proportion of participants using hypnotic medications for their sleep difficulties.
RESULTS
Self-Reported Sleep
Group comparisons. Significant group-time interactions were obtained on all sleep variables, with the exception of total sleep time: sleep efficiency (F1,52 = 22.59; P < .0001), total wake time (F1,52 = 22.77; P < .001), sleep onset latency F1,53 = 4.16; P < .05), wake after sleep onset (F1,52 = 16.70; P < .001), ISI-P (F1,52 = 25.31; P < .0001), ISI-C (F1,52 = 79.37; P < .0001), and ISI-SO (F1,48 = 4.54; P < .05). A priori contrasts revealed significant time effects on all variables in the treatment condition and all variables with the exception of two in the control condition (sleep onset latency and wake after sleep onset). Significant time effects found in the control condition were always of a lower magnitude compared with those of the treatment condition. For instance, sleep efficiency increased from 69.5% to 84.4% at post-treatment in the experimental condition, whereas it increased only from 71.1% to 74.5% in the control condition during the waiting period (Table 2; Figs 3 and 4).
An analysis was conducted to investigate whether hypnotic use at pretreatment had a moderating role in the effect of CBT on subjective sleep measures at post-treatment. No significant hypnotic use-group-time interaction was found on any of these sleep variables (P from .28 to .93).
Pooled analyses. Significant differences from pre- to post-treatment were obtained for all sleep variables (all P < 0.01). Regarding the maintenance of these benefits over time, significant differences between the post-treatment and the follow-up evaluations were obtained on only two variables (total sleep time, F3,163 = 2.75; P < .05, and ISI-SO, F3,144 = 3.55; P < .05), and both were in the direction of further improvement at follow-up compared with post-treatment.
PSG
Group comparisons. Analyses on PSG data were conducted while controlling for the influence of some behaviors (ie, caffeine and alcohol intake, sleep medication use, day napping, and pain in the last 24 hours) that were measured before each night spent in the laboratory. The consumption of caffeine the preceding day was significantly associated with increased sleep onset latency, whereas day napping was associated with increased sleep efficiency. The use of a hypnotic medication and alcohol was not significantly associated with any of the PSG data (data not presented).
No significant group-time interaction was observed on any of the sleep variables. However, the analyses revealed a marginally significant group-time interaction for sleep-onset latency (F1,46 = 3.88; P = .06; from 22.4 minutes at pre-treatment to 14.2 minutes at post-treatment), and for total sleep time (F1,34 = 3.64; P = .06; from 367.1 to 373.8 minutes), both in the direction of an improvement.
Pooled analyses. Significant differences from pretreatment to post-treatment were observed for all sleep measures, except total sleep time: sleep efficiency (F1,62 = 9.92; P < .05), total wake time (F1,62 = 15.91; P < .001), sleep onset latency (F1,70 = 12.92; P < .001), wake after sleep onset (F1,61 = 6.37; P < .05). Regarding the maintenance of gains over time, none of the variables significantly changed from post-treatment to the 6-month follow-up assessment.
Hypnotic Medication Use
Group comparisons. A significant group-time interaction was obtained on the frequency of medicated nights per week (including benzodiazepines and the newer nonbenzodiazepine hypnotics; F1,49 = 5.02; P < .05). On average, treated patients decreased their weekly usage of hypnotic medications from 1.68 to 0.98 nights at post-treatment, whereas it increased from 1.84 to 2.07 nights in the control group during this time (Table 3). There was no significant group-time interaction effect with regard to the average nightly dose of medication used (in diazepam equivalents) or the proportion of participants using hypnotic medications (21 = 1.37; P = .24), despite the fact that the latter decreased from 37.0% to 21.7% in the treatment condition and remained stable in the control condition (43.3% to 44.8%).
Pooled analyses. A significant reduction of the frequency of medicated nights per week was obtained from pre- to post-treatment (F1,167 = 18.90; P < .0001), and no significant time effect was observed from post-treatment to follow-up (Table 3). Similarly, analyses revealed a significant reduction of the average nightly dosage of hypnotic medications used (in diazepam equivalent) from pre- to post-treatment (F1,159 = 6.85; P < .01), but no significant differences from post-treatment to the follow-up assessments. Regarding the proportion of patients using an hypnotic medication, analyses conducted on data of pre- and post-treatment and 3-month follow-up (data of 6- and 12-month follow-ups were not included because of a high proportion of missing data), revealed a significant time effect (22 = 9.28; P < .01).
Psychological Symptoms and Quality of Life
Group comparisons. Significant group-time interactions were obtained on scores of anxiety (F1,45 = 5.19; P < .05), depression (F1,48 = 4.14; P < .05), and global quality of life (F1,48 = 5.69; P < .05). A priori contrasts revealed significant time effects in the treatment condition on anxiety (F1,46 = 4.77; P < .05), depression (F1,49 = 9.03; P < .01), and the global quality-of-life scale (F1,48 = 16.27; P < .001), whereas no significant time effect was found on any variable in the control condition (Table 4).
Pooled analyses. Pooled data revealed significant differences between pre- and post-treatment on anxiety (F1,150 = 11.10; P < .001), depression (F1,146 = 11.87; P < .001), and fatigue (F1,158 = 11.70; P < .001), and the global quality-of-life scale (F1,159 = 15.63; P < .0001). No significant difference was detected between post-treatment and the follow-up evaluations on any of these variables.
Clinical Significance of Treatment Effects
Group comparisons. At post-treatment, 56.5% of treated patients obtained a sleep efficiency of 85% or greater, as assessed with the sleep diary, compared with only 20.7% of the control group. Similarly, 56.5%, 91.3%, and 39.1% of the patients in the treatment group scored lower than 8 on the ISI-P, ISI-C, and ISI-SO at post-treatment, respectively, compared with only 10.3%, 0%, and 17.2% in the control group. Finally, 76.2% of the treated patients obtained a sleep efficiency of 85% or greater at post-treatment as assessed with PSG compared with 65.5% of the control group after their waiting period.
Pooled data. Table 5 presents the number (and percentages) of both groups of patients pooled together who meet criteria for a clinical significant change at post-treatment and follow-up evaluations. Between 35.3% and 88.2% of the patients had a clinically significant change at post-treatment depending on the criterion used. These percentages consistently increased at follow-up evaluations, with 62.5% to 85.0% of the patients meeting criteria for a clinically significant change at the 12-month follow-up evaluation. Subjective sleep assessments (sleep diary, ISI) yielded higher clinically significant increases compared with objective assessments (PSG), and improvements were reliably judged more clinically meaningful by the clinician (ISI-C and degree of change) compared with the patients and the significant others.
DISCUSSION
The goal of this study was to assess the efficacy of a cognitive-behavioral treatment for chronic insomnia secondary to cancer. These results indicate that treated patients showed a significantly greater improvement of their sleep at post-treatment, as assessed using self-reported instruments (ie, sleep diary, ISI), compared with control patients after their waiting period, and that there were no significant differences on therapeutic gains obtained whether patients were hypnotic users or not at pretreatment. However, PSG data were not significantly more improved in treated patients compared with control patients. Pooled analyses uniformly revealed that sleep, both assessed subjectively and objectively, improved significantly from pre- to post-treatment, and that these improvements were well maintained, and even enhanced in some cases, up to 12 months after the intervention. Sleep improvements were also found to be clinically significant in a substantial proportion of patients. The treatment of insomnia was associated with several other positive findings including reduced use of sleep medication, decreased anxiety and depression, and improved global quality of life.
The short-term effects of the intervention found on self-reported sleep measures are salient. On average, treated patients increased their sleep efficiency from 69% to 84% at post-treatment. Considering that a sleep efficiency of 85% is typically used to distinguish clinical insomnia from normal sleep, this is a significant finding. Also noteworthy is the finding that the mean ISI scores, as evaluated by the patient and the clinician, were lower than 8 (the cutoff score typically used to define clinical insomnia32,34) in treated patients, whereas they remained well above this cutoff score (approximately 14) in the untreated patients. The proportion of patients with clinically significant change was on average 70% at the 12-month follow-up evaluation. Thus, although patients with chronic insomnia generally benefited from the intervention, not all became good sleepers after CBT. This is consistent with studies on primary insomnia showing that approximately 70% to 80% of the patients benefit from a psychological intervention.25,26
The maintenance, and in some cases enhancement, of treatment gains during the follow-up period on all variables measured is another important finding of this study. The capacity of CBT to produce long-enduring effects has been demonstrated in the primary insomnia context,48,49 and is so largely acknowledged that it is now considered the treatment of choice for chronic insomnia.50
Another clinically important finding is the reduction of sleep medication use in the treated patients, despite the fact that no specific recommendations or strategies were offered to participants regarding drug withdrawal during the course of the intervention. This finding, consistent with the literature on primary insomnia,24,51,52 suggests that offering alternative strategies to medication for coping with sleep difficulties is often sufficient to help patients reduce or even stop the use of hypnotic medications. Other positive outcomes associated with the treatment included reduction of psychological distress (ie, depression and anxiety) and improvement of overall quality of life, which also is in agreement with previous studies.15,16,23,24
Conversely, analyses conducted on PSG data are less convincing. Pooled analyses generally revealed significant sleep improvements from pre- to post-treatment, and maintenance of these gains during the follow-up period. However, only two variables (ie, sleep-onset latency and total sleep time) were more improved in the treated patients at post-treatment; these differences only approached significance. Although unexpected, this finding is consistent with another study conducted in geriatric patients with various medical illnesses that found no significant effect of CBT for insomnia on an objective sleep measure (ie, actigraphy).11 Findings of our study may be interpreted in several different ways. It may be that sleep improvements that occurred over time, as assessed with PSG, were not related to the treatment but simply to the passage of time or a greater adaptation to the sleep laboratory with time. Another explanation, which we think is more plausible, relates to the fact that patients were selected for the study on the basis of subjective sleep data only. A comparison of data presented in Table 2 reveals that patients had much better PSG indices compared with sleep diary indices at pretreatment, thus leaving little space for improvement with treatment as measured with PSG (because of a floor effect). Thus, it may be that these small treatment effects became only statistically significant with the pooled data set because of increased statistical power. A lack of correlation between subjective and objective sleep measures is common,53 and is often interpreted as a consequence of the weak ecologic validity of PSG, especially when patients spend only a few nights in the laboratory, rather than a lack of validity of subjective measures.54,55 Indeed, insomnia is a condition that is defined primarily by a subjective complaint of poor sleep and by a subjective dissatisfaction with sleep that are better accounted by sleep diaries and questionnaires.56
This study is characterized by several strengths (eg, randomization, operational definition of chronic and secondary insomnia, manualized treatment). In addition, patients with comorbid depressive and anxiety disorder and those using psychotropic medications were included in the study (under certain conditions), which maximizes the generalization of these findings. Conversely, the fact that participants were all white, were mostly well educated, were all breast cancer survivors, were recruited through ads (reflecting a high level of motivation), and constituted a relatively small proportion of all patients who were screened for the study may limit the generalization of the findings. The study is further limited by the use of a waiting-list control condition that did not control for nonspecific therapeutic ingredients. It is therefore impossible to determine whether sleep improvements observed in this study are really attributable to the cognitive-behavioral strategies used or to other ingredients common to all psychotherapeutic approaches (eg, therapist empathy, group support).
This study nonetheless has several important clinical implications. It emphasizes the importance of screening more systematically for clinical insomnia among patients with cancer, a problem that has largely been neglected in the oncologic practice.1 The study also highlights the importance of offering interventions specifically targeting insomnia on a more routine basis in this population. In addition, this study shows that a psychological intervention principally developed for primary insomnia is also efficacious for insomnia secondary to a serious physical illness, without altering significantly the main treatment components.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
Acknowledgment
The authors thank Catherine Quesnel, Isabelle Giguère, Lucie Casault, Séverine Hervouet, Véronique Dupéré, Aude Caplette-Gingras, Célyne Bastien, Manon Lamy, Annick Ferland, Chantal Kenney, Julie Dumont, Monique Hamel, and Pascal Lebreton for their important contributions.
NOTES
Supported in part by an operating grant (MT-14039) and salary support from the Canadian Institutes of Health Research.
Results of this study have been reported in part at the 16th Annual Meeting of the Associated Professional Sleep Societies, Seattle, WA, June 2002; the 62nd Annual Convention of the Canadian Psychological Association, Québec, Québec, Canada, June 2001; the 2nd Scientific Conference on Reasons for Hope: New Developments in Breast Cancer Research, Québec, Québec, Canada, May 2001; the 6th World Congress of Psycho-Oncology, Banff, Alberta, Canada, April 2003; and the 6th International Congress of Behavioral Medicine, Brisbane, Australia, November 2000.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Savard J, Morin CM: Insomnia in the context of cancer: A review of a neglected problem. J Clin Oncol 19:895-908, 2001
Davidson JR, MacLean AW, Brundage MD, et al: Sleep disturbance in cancer patients. Soc Sci Med 54:1309-1321, 2002
Savard J, Simard S, Blanchet J, et al: Prevalence, clinical characteristics, and risk factors for insomnia in the context of breast cancer. Sleep 24:583-589, 2001
Simon GE, VonKorff M: Prevalence, burden, and treatment of insomnia in primary care. Am J Psychiatry 154:1417-1423, 1997
Hall N: Taking policy action to reduce benzodiazepine use and promote self-care among seniors. J Appl Gerontol 17:318-351, 1998
National Institutes of Health Consensus Conference: Drugs and insomnia—The use of medication to promote sleep. JAMA 251:2410-2414, 1984
National Institutes of Health: Consensus development conference statement: The treatment of sleep disorders of older people. Sleep 14:169-177, 1991
National Institutes of Health: NIH releases statement on behavioral and relaxation approaches for chronic pain and insomnia. Am Fam Physician 53:1877-1880, 1996
Currie SR, Wilson KG, Pontefract AJ, et al: Cognitive-behavioral treatment of insomnia secondary to chronic pain. J Consult Clin Psychol 68:407-416, 2000
Lichstein KL, Wilson NM, Johnson CT: Psychological treatment of secondary insomnia. Psychol Aging 15:232-240, 2000
Rybarczyk B, Lopez M, Benson R, et al: Efficacy of two behavioral treatment programs for comorbid geriatric insomnia. Psychol Aging 17:288-298, 2002
Cannici J, Malcom R, Peek LA: Treatment of insomnia in cancer patients using muscle relaxation training. J Behav Ther Exp Psychiatry 14:251-256, 1983
Stam HJ, Bultz BD: The treatment of severe insomnia in a cancer patient. J Behav Ther Exp Psychiatry 17:33-37, 1986
Simeit R, Deck R, Conta-Marx B: Sleep management training for cancer patients with insomnia. Support Care Cancer 12:176-183, 2004
Davidson JR, Waisberg JL, Brundage MD, et al: Nonpharmacologic group treatment of insomnia: A preliminary study with cancer survivors. Psychooncology 10:389-397, 2001
Quesnel C, Savard J, Simard S, et al: Efficacy of cognitive-behavioral therapy for insomnia in women treated for nonmetastatic breast cancer. J Consult Clin Psychol 71:189-200, 2003
Savard J, Simard S, Ivers H, et al: Randomized study on the efficacy of cognitive-behavioral therapy for insomnia secondary to breast cancer, Part II: Immunologic effects. J Clin Oncol 23:6097-6106, 2005
American Sleep Disorders Association: International Classification of Sleep Disorders (ICSD): Diagnostic and Coding Manual. Rochester, MN, American Sleep Disorders Association, 1997
American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders (ed 4). Washington, DC, American Psychiatric Association, 1994
Morin CM: Insomnia: Psychological Assessment and Management. New York, The Guilford Press, 1993
Derogatis LR, Feldstein M, Morrow G, et al: A survey of psychotropic drug prescriptions in an oncology population. Cancer 44:1919-1929, 1979
Stiefel FC, Kornblith AB, Holland JC: Changes in the prescription patterns of psychotropic drugs for cancer patients during a 10-year period. Cancer 65:1048-1053, 1990
Backhaus J, Hohagen F, Voderholzer U, et al: Long-term effectiveness of a short-term cognitive-behavioral group treatment for primary insomnia. Eur Arch Psychiatry Clin Neurosci 251:35-41, 2001
Espie CA, Inglis SJ, Tessier S, et al: The clinical effectiveness of cognitive behaviour therapy for chronic insomnia: Implementation and evaluation of a sleep clinic in general medical practice. Behav Res Ther 39:45-60, 2001
Murtagh DRR, Greenwood KM: Identifying effective psychological treatments for insomnia: A meta-analysis. J Consult Clin Psychol 63:79-89, 1995
Morin CM, Culbert JP, Schwartz SM: Nonpharmacological interventions for insomnia: A meta-analysis of treatment efficacy. Am J Psychiatry 151:1172-1180, 1994
McCrae CS, Lichstein KL: Secondary insomnia: Diagnostic challenges and intervention opportunities. Sleep Med Rev 5:47-61, 2001
McCrae CS, Lichstein KL: Secondary insomnia: A heuristic model and behavioral approaches to assessment, treatment, and prevention. Appl Prev Psychol 10:107-123, 2001
First MB, Spitzer RL, Gibbon M, et al: Structured Clinical Interview for DSM-IV Axis I Disorders: Patient Edition (SCID-I/P, Version 2.0). New York, NY, Biometrics Research Department, New York State Psychiatric Institute, 1996
Coates TJ, Killen JD, George J, et al: Estimating sleep parameters: A multitrait-multimethod analysis. J Consult Clin Psychol 50:345-352, 1982
Rechtschaffen A, Kales A: A Manual of Standardized Terminology: Techniques and Scoring System for Stages in Human Subjects. Bethesda, MD, US Dept of Health, Education, and Welfare, Public Health Service—National Institutes of Neurological Diseases and Blindness, Neurological Information Network, 1968
Savard MH, Savard J, Simard S, et al: Empirical validation of the Insomnia Severity Index in cancer patients. Psychooncology 14:429-441, 2005
Bastien CH, Vallieres A, Morin CM: Validation of the Insomnia Severity Index as an outcome measure for insomnia research. Sleep Med 2:297-307, 2001
Smith S, Trinder J: Detecting insomnia: Comparison of four self-report measures of sleep in a young adult population. J Sleep Res 10:229-235, 2001
Zigmond AS, Snaith RP: The Hospital Anxiety and Depression Scale. Acta Psychiatr Scand 67:361-370, 1983
Roth AJ, Kornblith AB, Batel-Copel L, et al: Rapid screening for psychologic distress in men with prostate carcinoma. Cancer 82:1904-1908, 1998
Savard J, Laberge B, Gauthier JG, et al: Screening clinical depression in HIV-seropositive patients using the Hospital Anxiety and Depression Scale. AIDS Behav 3:167-175, 1999
Savard J, Laberge B, Gauthier JG, et al: Evaluating anxiety and depression in HIV-infected patients. J Pers Assess 71:349-367, 1998
Smets EMA, Garssen B, Bonke B, et al: The multidimensional fatigue inventory (MFI) psychometric qualities of an instrument to assess fatigue. J Psychosom Res 39:315-325, 1995
Fillion L, Gelinas C, Simard S, et al: Validation evidence for the French Canadian adaptation of the Multidimensional Fatigue Inventory as a measure of cancer-related fatigue. Cancer Nurs 26:143-154, 2003
Aaronson NK, Ahmedzai S, Bergman B: The EORTC QLQ-C30: A quality of life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 85:365-376, 1993
Osoba D, Aaronson N, Zee B, et al: Modification of the EORTC QLQ-C30 (version 2.0) based on content validity and reliability testing in large samples of patients with cancer. Qual Life Res 6:103-108, 1997
Tabachnik BG, Fidell LS: Using Multivariate Statistics (ed 4). New York, NY, Harper Collins, 2001
SAS Institute: SAS/STAT User's Guide, Version 8: Volumes 1, 2, and 3. Cary, NC, SAS Institute, 2001
Baron RM, Kenny DA: The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. J Pers Soc Psychol 51:1173-1183, 1986
Keselman HJ, Algina J, Kowalchuk RK: The analysis of repeated measures designs: A review. Br J Math Stat Psych 54:1-20, 2001
Frigon J-Y, Laurencelle L: Analysis of covariance: A proposed algorithm. Educ Psychol Meas 53:1-18, 1993
McClusky HY, Milby JB, Switzer PK, et al: Efficacy of behavioral versus triazolam treatment in persistent sleep-onset insomnia. Am J Psychiatry 148:121-126, 1991
Morin CM, Colecchi C, Stone J, et al: Behavioral and pharmacological therapies for late-life insomnia: A randomized controlled trial. JAMA 281:991-999, 1999
Morin CM, Hauri PJ, Espie CA, et al: Nonpharmacologic treatment of chronic insomnia. Sleep 22:1134-1156, 1999
Jacobs GD, Benson H, Friedman R: Perceived benefits in a behavioral-medicine insomnia program: A clinical report. Am J Med 100:212-216, 1996
Morin CM, Kowatch RA, Barry T, et al: Cognitive-behavior therapy for late-life insomnia. J Consult Clin Psychol 61:137-146, 1993
Trinder J: Subjective insomnia without objective findings: A pseudo diagnostic classification? Psychol Bull 103:87-94, 1988
Morin CM: The nature of insomnia and the need to refine our diagnostic criteria. Psychosom Med 62:483-485, 2000
Wohlgemuth WK, Edinger JD, Fins AI, et al: How many nights are enough? The short-term stability of sleep parameters in elderly insomniacs and normal sleepers. Psychophysiology 36:233-244, 1999
Krystal AD: The changing perspective on chronic insomnia management. J Clin Psychiatry 65:20-25, 2004(Josée Savard, Sébastien S)
School of Psychology, Université Laval, Québec, Québec, Canada
ABSTRACT
PURPOSE: Chronic insomnia is highly prevalent in cancer patients. Cognitive-behavioral therapy (CBT) is considered the treatment of choice for chronic primary insomnia. However, no randomized controlled study has been conducted on its efficacy for insomnia secondary to cancer. Using a randomized controlled design, this study conducted among breast cancer survivors evaluated the effect of CBT on sleep, assessed both subjectively and objectively, and on hypnotic medication use, psychological distress, and quality of life.
PATIENTS AND METHODS: Fifty-seven women with insomnia caused or aggravated by breast cancer were randomly assigned to CBT (n = 27) or a waiting-list control condition (n = 30). The treatment consisted of eight weekly sessions administered in a group and combined the use of stimulus control, sleep restriction, cognitive therapy, sleep hygiene, and fatigue management. Follow-up evaluations were carried out 3, 6, and 12 months after the treatment.
RESULTS: Participants who received the insomnia treatment had significantly better subjective sleep indices (daily sleep diary, Insomnia Severity Index), a lower frequency of medicated nights, lower levels of depression and anxiety, and greater global quality of life at post-treatment compared with participants of the control group after their waiting period. Results were more equivocal on polysomnographic indices. Therapeutic effects were well maintained up to 12 months after the intervention and generally were clinically significant.
CONCLUSION: This study supports the efficacy of CBT for insomnia secondary to breast cancer.
INTRODUCTION
Insomnia occurs frequently in the context of cancer, affecting 30% to 50% of cancer patients.1 Insomnia is particularly frequent in breast cancer patients relative to other cancer types.2 A recent survey3 revealed that 51% of women treated for breast cancer reported nonspecific sleep difficulties and 19% met diagnostic criteria for an insomnia syndrome, which was chronic in 95% of the patients (insomnia duration of 6 months or more). Insomnia is associated with significant burden, both for the individual and for the health care system, in terms of functional impairments, reduced quality of life, and increased health care use,4 hence the importance of treating insomnia, particularly when it is chronic.
Psychological interventions should be the preferred alternative, compared with hypnotic medications, when insomnia is chronic.5-8 The efficacy of psychological interventions, especially cognitive-behavioral therapy (CBT), is well established for primary insomnia (ie, insomnia not etiologically related to a medical or a psychological condition). There is also recent evidence supporting the efficacy of CBT for insomnia secondary to chronic pain9 and to a medical or a psychological condition,10 as well as for insomnia comorbid with diverse medical conditions.11 However, studies conducted on insomnia associated with cancer are scarce.
Early pilot studies yielded mixed results on the efficacy of relaxation for insomnia in cancer patients.12,13 More recently, nonrandomized studies showed positive outcomes associated with multimodal psychological interventions for insomnia, including improved subjective sleep and some aspects of quality of life.14,15 We also found promising results in a pilot study using a multiple-baseline design, a more stringent definition of secondary insomnia, and objective sleep measures.16 This study conducted among eight women with insomnia caused or aggravated by breast cancer showed reduced total wake time and increased sleep efficiency, both assessed subjectively (ie, sleep diary) and objectively (ie, polysomnography), as well as improved mood, fatigue, and quality of life. However, a randomized trial is still needed to document more systematically the efficacy of CBT for insomnia secondary to cancer.
The goals of this study were to assess the efficacy of CBT for insomnia secondary to breast cancer, in a randomized controlled group design, on measures of subjective and objective sleep, psychological functioning (ie, depression, anxiety, fatigue), and quality of life. The study also investigated the effect of CBT for insomnia on immunologic functioning. These results will be reported in a companion article because of space limitations (Savard et al17).
PATIENTS AND METHODS
Participants
The participants were recruited between November 1998 and December 2001 with fliers and pamphlets, ads placed in the local newspapers, and by physician referrals. To be accepted in the study, participants had to have completed radiotherapy and chemotherapy for a stage I to III breast cancer at least 1 month prior to enrollment onto the study, and had to meet diagnostic criteria for a chronic insomnia syndrome, as defined by the combined criteria of the International Classification of Sleep Disorders18 and of the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition (DSM-IV),19 as well as those commonly used in insomnia research.20 Those criteria included difficulty in initiating and/or maintaining sleep, whereby sleep-onset latency and/or wake after sleep onset is greater than 30 minutes; sleep efficiency (ratio of total sleep time to total time spent in bed) lower than 85%; difficulties occurring at least 3 nights per week; difficulties occurring for at least 6 months; and difficulties causing marked distress or significant impairment in daytime functioning (eg, fatigue, disturbed mood, performance deficits). Only patients whose insomnia was judged to be secondary to cancer were included in the study (ie, those whose sleep difficulties were caused or aggravated by the cancer diagnosis or treatment).
The following exclusion criteria were also used: presence of severe major depression or another serious psychiatric disorder; presence of a sleep disorder other than insomnia (eg, sleep apnea, periodic limb movements); presence of another illness affecting the immune system (eg, HIV infection); regular use of a psychotropic medication other than hypnotics (eg, antidepressants), unless the dosage used was stable in the last month and did not increase during the study; and current involvement in psychotherapy.
Patients using a stable dosage of hypnotic medication were not excluded from the study, provided that they met all criteria described, including the diagnostic criteria for an insomnia syndrome. Thus, hypnotic users who were included in the study were all reporting clinically significant insomnia despite the use of this medication (hence suggesting its limited utility). Including those patients has the advantage of increasing the study's external validity. Indeed, it has been found that 20% to 40% of cancer patients use a psychotropic medication for sleep difficulties.2,21,22 This variable (changes in hypnotic use) was used as an outcome measure in this study. Several other studies conducted in the context of primary23,24 or secondary9,11 insomnia have included patients using a hypnotic medication and used this variable as an outcome measure. Statistical analyses were nonetheless performed to ensure the inclusion of these patients did not influence the effect of the intervention on other sleep variables (see Statistical Analyses).
Sample characteristics. A priori power analyses, based on effect sizes obtained from two meta-analyses assessing the efficacy of psychological interventions for primary insomnia,25,26 yielded a total sample size estimate for this study ranging from 50 to 60 participants. Figure 1 shows rates of recruitment, exclusion, refusal, and dropout throughout the study. As illustrated, of the 160 women who responded to the recruitment effort, 58 (36.3% of women who were initially screened over the phone) were enrolled in the study and randomly assigned to the treatment or the control condition. Demographic and clinical characteristics of the final sample of 57 participants at study entry are listed in Table 1. All patients were white. Of all demographic and clinical variables examined, the only group difference at pretreatment was for the proportion of patients with a comorbid physical illness (eg, cardiovascular disease, arthritis), which was greater in the treatment (59%) than in the control condition (27%; 21 = 6.19; P < .05).
Experimental Design
Participants were randomly assigned to one of the two following conditions: CBT administered immediately or waiting-list control condition (WLC; Fig 2). Because the recruitment was somewhat slow and treatment was conducted in small groups, the random allocation of patients was done by blocks of five to six patients. Participants assigned to the control group waited a minimum of 8 weeks, which corresponded to the duration of the intervention, were assessed again on study variables, including completion of a sleep diary for a 2-week period, and then received CBT. This second assessment (pretreatment or postwaiting assessment) of control patients was contrasted with the post-treatment evaluation of treated patients to verify the short-term effect of CBT. Additional evaluations were conducted 3, 6, and 12 months after the end of their respective treatment to assess the maintenance of treatment effects over time. The study was approved by the ethical review boards of L'H?tel-Dieu de Québec and Université Laval (Québec, Canada).
Measures
Insomnia Interview Schedule. The Insomnia Interview Schedule (IIS)20 consists of a series of semistructured questions evaluating several aspects of sleep difficulties. It also provides guidelines for diagnosing insomnia according to the criteria described and for detecting symptoms suggestive of other sleep disorders (eg, obstructive sleep apnea, periodic limb movement disorder). For the purpose of this study, questions were added to the IIS to determine whether insomnia was secondary to breast cancer by assessing the temporal relationship between the cancer diagnosis (or treatment) and the onset or aggravation of sleep disturbances.10,27,28
Structured Clinical Interview for the DSM-IV. The Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (SCID),29 is used to evaluate the presence of current and past psychiatric disorders according to DSM-IV diagnostic criteria.19
Sleep diary. The sleep diary20 (subjective sleep measure) was completed each day on arising and provided information pertaining to the use of hypnotic medications as sleep aids, bedtime hour, time of falling asleep, number and duration of awakenings during the night, and the time of morning awakening and arising from bed. The dependent variables derived from that information were sleep onset latency (time to sleep after lights out), wake after sleep onset (summation of nocturnal awakenings), total wake time (summation of sleep onset latency, wake after sleep onset, and early morning awakening), total sleep time (time in bed minus total wake time), sleep efficiency (ratio of total sleep time to the actual time spent in bed multiplied by 100), and use of sleep-promoting medications (type, frequency, and quantity). Although subjective reports do not always correspond to polysomnographic sleep data, they still provide a reliable index of insomnia.30
Polysomnography. The polysomnography (PSG; objective sleep measure) montage included standard electroencephalographic, electromyographic, and electro-oculographic monitoring. On the first night in the laboratory, respiration (airflow, tidal volume, and oxygen saturation) and anterior tibialis electromyography were also measured to rule out the presence of sleep apnea or periodic limb movements. The sleep stages were scored in 30-second epochs, using standard criteria31 by an experienced technician who was blinded to the patient's experimental condition. The following dependent variables were derived from the polysomnographic data: sleep onset latency (time from lights out to persistent sleep), wake after sleep onset (total of nocturnal awakenings), total wake time (latency to sleep + wake after sleep onset + early morning awakening), total sleep time (time in bed minus total wake time), and sleep efficiency (ratio of total sleep time to the actual time spent in bed multiplied by 100). These sleep variables were averaged from nights 2 and 3 for each study time assessment, with night 1 excluded to allow for patients' adaptation to the laboratory setting.
Insomnia Severity Index. The Insomnia Severity Index (ISI)20 is a seven-item questionnaire designed to evaluate insomnia severity on the basis of difficulties falling asleep, night-time awakenings, early morning awakenings, impairment of daytime functioning due to sleep problems, noticeability of impairments, distress or worry caused by sleep difficulties, and dissatisfaction with sleep. Each item is rated using a five-point Likert scale ranging from 0 (not at all) to 4 (very much), for a total score ranging from 0 to 28. In this study, three parallel versions of the ISI were administered: one that was completed by the participant (ISI-P), one that was completed by the clinician (ISI-C), and one that was completed by a participant's significant other (ISI-SO). We recently validated the French-Canadian version of the ISI specifically in the context of cancer,32 and found it had psychometric properties equivalent to those in the original version.33,34
Hospital Anxiety and Depression Scale. This questionnaire includes 14 items divided into two subscales: depression (Hospital Anxiety and Depression Scale [HADS] -D35: seven items) and anxiety (HADS-A: seven items). A score of 7 and more, for a maximum score of 21, at each subscale suggests the presence of clinical levels of depression or anxiety.36,37 The French-Canadian version proved to possess psychometric qualities equivalent to those of the original English version.38
Multidimensional Fatigue Inventory. The French-Canadian version of the Multidimensional Fatigue Inventory (MFI),39 which possesses adequate psychometric properties,40 is a short form of the original MFI. It contains 15 items (on a scale from 1 to 5) divided into four subscales including general and physical fatigue, reduction in activities, reduction in motivation, and mental fatigue. A global score of fatigue is obtained by calculating a mean for all items.
The European Organization for Research and Treatment of Cancer Quality of Life Questionnaire. The European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (QLQ-C30+3)41,42 was developed and validated with cancer patients. Only results obtained on the global quality-of-life scale, comprised of three items, are reported in this article. Scores are transformed to give a score ranging from 0 to 100. The French version was developed by the authors of the original English version.
Procedure
Telephone screening. A telephone screening was undertaken with women seeking information about the study to evaluate their eligibility briefly and explain the study goals and procedures. A clinical interview was scheduled with interested and eligible patients based on initial criteria (eg, sleep difficulties at least 3 nights a week for at least 6 months).
Clinical interview. During this interview, study procedures were explained in detail to participants and their written consent was obtained. Then, the IIS and the SCID were administered individually by two trained and independent clinicians. The clinician who administered the IIS also completed the ISI-C after the interview. After the evaluations, the two clinicians met to discuss the information obtained and to decide on the eligibility of the participant. Eligible participants were asked to complete a daily sleep diary for a period of 2 weeks. At this point, patients were considered eligible for the study when information of sleep diaries confirmed data obtained with the IIS concerning the frequency (ie, at least 3 nights per week) and severity (ie, awakenings > 30 minutes, sleep efficiency < 85%) of sleep difficulties.
PSG. After the clinical interview, participants spent 3 consecutive nights in a sleep laboratory for PSG assessment. The day after, patients with no evidence of sleep apnea or periodic limb movement based on PSG were definitely accepted onto the study (which was the case for all patients) and randomly assigned to treatment or control conditions. Patients allocated to the control group completed the battery of self-report scales (ISI-P, HADS, MFI, and QLQ-C30+3) and were instructed to ask a significant other to complete the ISI-SO, which provided the prewaiting measures.
Information session. Before beginning the intervention, participants were invited to an information session explaining the general treatment procedures. All participants also had to complete the same battery of self-report scales, which provided the pretreatment measures, and to ask a significant other to complete the ISI-SO.
Treatment. The insomnia treatment consisted of eight weekly sessions of approximately 90 minutes, offered in groups of four to six patients. The treatment protocol was based on clinical procedures developed by Morin20 and slightly adapted by our team for the cancer population. This multimodal approach combined behavioral (ie, stimulus control therapy, sleep restriction), cognitive (ie, cognitive restructuring), and educational (ie, sleep hygiene, fatigue and stress management) strategies that were described in a treatment manual given to all participants. Participants were instructed to continue completing the daily sleep diary throughout the treatment for clinical purposes. The treatment content is described in more detail elsewhere.16
The treatment was administered by a master-level psychologist with experience in the administration of this particular treatment protocol. An optional booster session was offered to participants 1 month after the end of the treatment. Participants interested in reducing their use of sleep-promoting medication were advised to consult a physician or pharmacist for supervision of medication withdrawal. Because missed treatment sessions (n = 8) were rescheduled, all patients received the entire treatment program, excluding the four patients who dropped out of the study during the course of the intervention (Fig 1).
Post-treatment procedure and follow-up. After the intervention, as well as 3, 6, and 12 months after the end of treatment, the participants again completed the battery of self-report scales and completed daily sleep diaries for another 2-week period. The clinician and the significant other also rated their respective versions of the ISI at these times. In addition, at post-treatment and 6-month follow-up, the participants spent 3 consecutive nights in the laboratory to measure therapeutic progress objectively.
Statistical Analyses
All data were carefully inspected to identify missing data and outliers, and to assess normality.43 Descriptive and inferential statistics were completed using SAS 8.2 statistical software (SAS Institute, Cary, NC).44 For all inferential tests, = .05 (two tailed). The main analyses were based on a split-plot group (two conditions) -time (five assessments: pretreatment, post-treatment, and 3-, 6-, and 12-month follow-up) randomized design. Two subsets of analyses were performed. First, analyses were conducted to determine whether treated patients had greater improvements on all dependent variables at post-treatment compared with patients in the control group after their waiting period. These findings are reported in Group Comparisons; because of this objective, only significant group-time interactions are reported. Then, because of the nonsignificant group effects obtained, data of both groups were pooled together to confirm, with a larger sample size, the benefits associated with the intervention at post-treatment, and to evaluate whether therapeutic gains observed at post-treatment were maintained at follow-up assessments. These findings are reported in Pooled Analyses and significant time effects are emphasized. A secondary analysis was also performed to investigate the moderating role of hypnotic use at pretreatment on the effect of CBT on subjective sleep measures (sleep diary and ISI). This was done by testing whether the third-order interaction (hypnotic use-group-time) was significant.45
Data were analyzed within an intent-to-treat framework. Linear mixed models were used to test group, time, and group-time interaction effects for all continuous dependent variables. A priori contrasts were used to break down these effects. Satterthwaite F tests were computed because they are typically more robust to non-normality, unbalanced data, and violations of multisample sphericity.46
In accordance with the strategy suggested by Frigon and Laurencelle,47 various covariates (eg, hormone therapy and the presence of a comorbid medical condition) were tested to assess their capacity to reduce the error term. To be included in the mixed-model analyses as a covariate, a variable had to meet these two criteria: significant between-group differences and significant reduction of error variance of more than one dependent variable. As mentioned, significant between-group differences were found on only one variable—the presence of a comorbid illness—but this variable was only significantly associated with total sleep time as assessed with the sleep diary. Hence, results reported thereafter are those based on analyses conducted without covariates. For analyses conducted on PSG data, however, five covariates were added to the mixed models to control for the potential impact on sleep of various health behaviors the patients had during the 24 hours preceding each recording night (ie, caffeine and alcohol consumption, sleep medication use, napping, and pain level).
Finally, generalized estimating equations (SAS CATMOD procedure; SAS Institute) analyses were used to estimate the three main effects (group, time, and group-time interaction) for the proportion of participants using hypnotic medications for their sleep difficulties.
RESULTS
Self-Reported Sleep
Group comparisons. Significant group-time interactions were obtained on all sleep variables, with the exception of total sleep time: sleep efficiency (F1,52 = 22.59; P < .0001), total wake time (F1,52 = 22.77; P < .001), sleep onset latency F1,53 = 4.16; P < .05), wake after sleep onset (F1,52 = 16.70; P < .001), ISI-P (F1,52 = 25.31; P < .0001), ISI-C (F1,52 = 79.37; P < .0001), and ISI-SO (F1,48 = 4.54; P < .05). A priori contrasts revealed significant time effects on all variables in the treatment condition and all variables with the exception of two in the control condition (sleep onset latency and wake after sleep onset). Significant time effects found in the control condition were always of a lower magnitude compared with those of the treatment condition. For instance, sleep efficiency increased from 69.5% to 84.4% at post-treatment in the experimental condition, whereas it increased only from 71.1% to 74.5% in the control condition during the waiting period (Table 2; Figs 3 and 4).
An analysis was conducted to investigate whether hypnotic use at pretreatment had a moderating role in the effect of CBT on subjective sleep measures at post-treatment. No significant hypnotic use-group-time interaction was found on any of these sleep variables (P from .28 to .93).
Pooled analyses. Significant differences from pre- to post-treatment were obtained for all sleep variables (all P < 0.01). Regarding the maintenance of these benefits over time, significant differences between the post-treatment and the follow-up evaluations were obtained on only two variables (total sleep time, F3,163 = 2.75; P < .05, and ISI-SO, F3,144 = 3.55; P < .05), and both were in the direction of further improvement at follow-up compared with post-treatment.
PSG
Group comparisons. Analyses on PSG data were conducted while controlling for the influence of some behaviors (ie, caffeine and alcohol intake, sleep medication use, day napping, and pain in the last 24 hours) that were measured before each night spent in the laboratory. The consumption of caffeine the preceding day was significantly associated with increased sleep onset latency, whereas day napping was associated with increased sleep efficiency. The use of a hypnotic medication and alcohol was not significantly associated with any of the PSG data (data not presented).
No significant group-time interaction was observed on any of the sleep variables. However, the analyses revealed a marginally significant group-time interaction for sleep-onset latency (F1,46 = 3.88; P = .06; from 22.4 minutes at pre-treatment to 14.2 minutes at post-treatment), and for total sleep time (F1,34 = 3.64; P = .06; from 367.1 to 373.8 minutes), both in the direction of an improvement.
Pooled analyses. Significant differences from pretreatment to post-treatment were observed for all sleep measures, except total sleep time: sleep efficiency (F1,62 = 9.92; P < .05), total wake time (F1,62 = 15.91; P < .001), sleep onset latency (F1,70 = 12.92; P < .001), wake after sleep onset (F1,61 = 6.37; P < .05). Regarding the maintenance of gains over time, none of the variables significantly changed from post-treatment to the 6-month follow-up assessment.
Hypnotic Medication Use
Group comparisons. A significant group-time interaction was obtained on the frequency of medicated nights per week (including benzodiazepines and the newer nonbenzodiazepine hypnotics; F1,49 = 5.02; P < .05). On average, treated patients decreased their weekly usage of hypnotic medications from 1.68 to 0.98 nights at post-treatment, whereas it increased from 1.84 to 2.07 nights in the control group during this time (Table 3). There was no significant group-time interaction effect with regard to the average nightly dose of medication used (in diazepam equivalents) or the proportion of participants using hypnotic medications (21 = 1.37; P = .24), despite the fact that the latter decreased from 37.0% to 21.7% in the treatment condition and remained stable in the control condition (43.3% to 44.8%).
Pooled analyses. A significant reduction of the frequency of medicated nights per week was obtained from pre- to post-treatment (F1,167 = 18.90; P < .0001), and no significant time effect was observed from post-treatment to follow-up (Table 3). Similarly, analyses revealed a significant reduction of the average nightly dosage of hypnotic medications used (in diazepam equivalent) from pre- to post-treatment (F1,159 = 6.85; P < .01), but no significant differences from post-treatment to the follow-up assessments. Regarding the proportion of patients using an hypnotic medication, analyses conducted on data of pre- and post-treatment and 3-month follow-up (data of 6- and 12-month follow-ups were not included because of a high proportion of missing data), revealed a significant time effect (22 = 9.28; P < .01).
Psychological Symptoms and Quality of Life
Group comparisons. Significant group-time interactions were obtained on scores of anxiety (F1,45 = 5.19; P < .05), depression (F1,48 = 4.14; P < .05), and global quality of life (F1,48 = 5.69; P < .05). A priori contrasts revealed significant time effects in the treatment condition on anxiety (F1,46 = 4.77; P < .05), depression (F1,49 = 9.03; P < .01), and the global quality-of-life scale (F1,48 = 16.27; P < .001), whereas no significant time effect was found on any variable in the control condition (Table 4).
Pooled analyses. Pooled data revealed significant differences between pre- and post-treatment on anxiety (F1,150 = 11.10; P < .001), depression (F1,146 = 11.87; P < .001), and fatigue (F1,158 = 11.70; P < .001), and the global quality-of-life scale (F1,159 = 15.63; P < .0001). No significant difference was detected between post-treatment and the follow-up evaluations on any of these variables.
Clinical Significance of Treatment Effects
Group comparisons. At post-treatment, 56.5% of treated patients obtained a sleep efficiency of 85% or greater, as assessed with the sleep diary, compared with only 20.7% of the control group. Similarly, 56.5%, 91.3%, and 39.1% of the patients in the treatment group scored lower than 8 on the ISI-P, ISI-C, and ISI-SO at post-treatment, respectively, compared with only 10.3%, 0%, and 17.2% in the control group. Finally, 76.2% of the treated patients obtained a sleep efficiency of 85% or greater at post-treatment as assessed with PSG compared with 65.5% of the control group after their waiting period.
Pooled data. Table 5 presents the number (and percentages) of both groups of patients pooled together who meet criteria for a clinical significant change at post-treatment and follow-up evaluations. Between 35.3% and 88.2% of the patients had a clinically significant change at post-treatment depending on the criterion used. These percentages consistently increased at follow-up evaluations, with 62.5% to 85.0% of the patients meeting criteria for a clinically significant change at the 12-month follow-up evaluation. Subjective sleep assessments (sleep diary, ISI) yielded higher clinically significant increases compared with objective assessments (PSG), and improvements were reliably judged more clinically meaningful by the clinician (ISI-C and degree of change) compared with the patients and the significant others.
DISCUSSION
The goal of this study was to assess the efficacy of a cognitive-behavioral treatment for chronic insomnia secondary to cancer. These results indicate that treated patients showed a significantly greater improvement of their sleep at post-treatment, as assessed using self-reported instruments (ie, sleep diary, ISI), compared with control patients after their waiting period, and that there were no significant differences on therapeutic gains obtained whether patients were hypnotic users or not at pretreatment. However, PSG data were not significantly more improved in treated patients compared with control patients. Pooled analyses uniformly revealed that sleep, both assessed subjectively and objectively, improved significantly from pre- to post-treatment, and that these improvements were well maintained, and even enhanced in some cases, up to 12 months after the intervention. Sleep improvements were also found to be clinically significant in a substantial proportion of patients. The treatment of insomnia was associated with several other positive findings including reduced use of sleep medication, decreased anxiety and depression, and improved global quality of life.
The short-term effects of the intervention found on self-reported sleep measures are salient. On average, treated patients increased their sleep efficiency from 69% to 84% at post-treatment. Considering that a sleep efficiency of 85% is typically used to distinguish clinical insomnia from normal sleep, this is a significant finding. Also noteworthy is the finding that the mean ISI scores, as evaluated by the patient and the clinician, were lower than 8 (the cutoff score typically used to define clinical insomnia32,34) in treated patients, whereas they remained well above this cutoff score (approximately 14) in the untreated patients. The proportion of patients with clinically significant change was on average 70% at the 12-month follow-up evaluation. Thus, although patients with chronic insomnia generally benefited from the intervention, not all became good sleepers after CBT. This is consistent with studies on primary insomnia showing that approximately 70% to 80% of the patients benefit from a psychological intervention.25,26
The maintenance, and in some cases enhancement, of treatment gains during the follow-up period on all variables measured is another important finding of this study. The capacity of CBT to produce long-enduring effects has been demonstrated in the primary insomnia context,48,49 and is so largely acknowledged that it is now considered the treatment of choice for chronic insomnia.50
Another clinically important finding is the reduction of sleep medication use in the treated patients, despite the fact that no specific recommendations or strategies were offered to participants regarding drug withdrawal during the course of the intervention. This finding, consistent with the literature on primary insomnia,24,51,52 suggests that offering alternative strategies to medication for coping with sleep difficulties is often sufficient to help patients reduce or even stop the use of hypnotic medications. Other positive outcomes associated with the treatment included reduction of psychological distress (ie, depression and anxiety) and improvement of overall quality of life, which also is in agreement with previous studies.15,16,23,24
Conversely, analyses conducted on PSG data are less convincing. Pooled analyses generally revealed significant sleep improvements from pre- to post-treatment, and maintenance of these gains during the follow-up period. However, only two variables (ie, sleep-onset latency and total sleep time) were more improved in the treated patients at post-treatment; these differences only approached significance. Although unexpected, this finding is consistent with another study conducted in geriatric patients with various medical illnesses that found no significant effect of CBT for insomnia on an objective sleep measure (ie, actigraphy).11 Findings of our study may be interpreted in several different ways. It may be that sleep improvements that occurred over time, as assessed with PSG, were not related to the treatment but simply to the passage of time or a greater adaptation to the sleep laboratory with time. Another explanation, which we think is more plausible, relates to the fact that patients were selected for the study on the basis of subjective sleep data only. A comparison of data presented in Table 2 reveals that patients had much better PSG indices compared with sleep diary indices at pretreatment, thus leaving little space for improvement with treatment as measured with PSG (because of a floor effect). Thus, it may be that these small treatment effects became only statistically significant with the pooled data set because of increased statistical power. A lack of correlation between subjective and objective sleep measures is common,53 and is often interpreted as a consequence of the weak ecologic validity of PSG, especially when patients spend only a few nights in the laboratory, rather than a lack of validity of subjective measures.54,55 Indeed, insomnia is a condition that is defined primarily by a subjective complaint of poor sleep and by a subjective dissatisfaction with sleep that are better accounted by sleep diaries and questionnaires.56
This study is characterized by several strengths (eg, randomization, operational definition of chronic and secondary insomnia, manualized treatment). In addition, patients with comorbid depressive and anxiety disorder and those using psychotropic medications were included in the study (under certain conditions), which maximizes the generalization of these findings. Conversely, the fact that participants were all white, were mostly well educated, were all breast cancer survivors, were recruited through ads (reflecting a high level of motivation), and constituted a relatively small proportion of all patients who were screened for the study may limit the generalization of the findings. The study is further limited by the use of a waiting-list control condition that did not control for nonspecific therapeutic ingredients. It is therefore impossible to determine whether sleep improvements observed in this study are really attributable to the cognitive-behavioral strategies used or to other ingredients common to all psychotherapeutic approaches (eg, therapist empathy, group support).
This study nonetheless has several important clinical implications. It emphasizes the importance of screening more systematically for clinical insomnia among patients with cancer, a problem that has largely been neglected in the oncologic practice.1 The study also highlights the importance of offering interventions specifically targeting insomnia on a more routine basis in this population. In addition, this study shows that a psychological intervention principally developed for primary insomnia is also efficacious for insomnia secondary to a serious physical illness, without altering significantly the main treatment components.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
Acknowledgment
The authors thank Catherine Quesnel, Isabelle Giguère, Lucie Casault, Séverine Hervouet, Véronique Dupéré, Aude Caplette-Gingras, Célyne Bastien, Manon Lamy, Annick Ferland, Chantal Kenney, Julie Dumont, Monique Hamel, and Pascal Lebreton for their important contributions.
NOTES
Supported in part by an operating grant (MT-14039) and salary support from the Canadian Institutes of Health Research.
Results of this study have been reported in part at the 16th Annual Meeting of the Associated Professional Sleep Societies, Seattle, WA, June 2002; the 62nd Annual Convention of the Canadian Psychological Association, Québec, Québec, Canada, June 2001; the 2nd Scientific Conference on Reasons for Hope: New Developments in Breast Cancer Research, Québec, Québec, Canada, May 2001; the 6th World Congress of Psycho-Oncology, Banff, Alberta, Canada, April 2003; and the 6th International Congress of Behavioral Medicine, Brisbane, Australia, November 2000.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Savard J, Morin CM: Insomnia in the context of cancer: A review of a neglected problem. J Clin Oncol 19:895-908, 2001
Davidson JR, MacLean AW, Brundage MD, et al: Sleep disturbance in cancer patients. Soc Sci Med 54:1309-1321, 2002
Savard J, Simard S, Blanchet J, et al: Prevalence, clinical characteristics, and risk factors for insomnia in the context of breast cancer. Sleep 24:583-589, 2001
Simon GE, VonKorff M: Prevalence, burden, and treatment of insomnia in primary care. Am J Psychiatry 154:1417-1423, 1997
Hall N: Taking policy action to reduce benzodiazepine use and promote self-care among seniors. J Appl Gerontol 17:318-351, 1998
National Institutes of Health Consensus Conference: Drugs and insomnia—The use of medication to promote sleep. JAMA 251:2410-2414, 1984
National Institutes of Health: Consensus development conference statement: The treatment of sleep disorders of older people. Sleep 14:169-177, 1991
National Institutes of Health: NIH releases statement on behavioral and relaxation approaches for chronic pain and insomnia. Am Fam Physician 53:1877-1880, 1996
Currie SR, Wilson KG, Pontefract AJ, et al: Cognitive-behavioral treatment of insomnia secondary to chronic pain. J Consult Clin Psychol 68:407-416, 2000
Lichstein KL, Wilson NM, Johnson CT: Psychological treatment of secondary insomnia. Psychol Aging 15:232-240, 2000
Rybarczyk B, Lopez M, Benson R, et al: Efficacy of two behavioral treatment programs for comorbid geriatric insomnia. Psychol Aging 17:288-298, 2002
Cannici J, Malcom R, Peek LA: Treatment of insomnia in cancer patients using muscle relaxation training. J Behav Ther Exp Psychiatry 14:251-256, 1983
Stam HJ, Bultz BD: The treatment of severe insomnia in a cancer patient. J Behav Ther Exp Psychiatry 17:33-37, 1986
Simeit R, Deck R, Conta-Marx B: Sleep management training for cancer patients with insomnia. Support Care Cancer 12:176-183, 2004
Davidson JR, Waisberg JL, Brundage MD, et al: Nonpharmacologic group treatment of insomnia: A preliminary study with cancer survivors. Psychooncology 10:389-397, 2001
Quesnel C, Savard J, Simard S, et al: Efficacy of cognitive-behavioral therapy for insomnia in women treated for nonmetastatic breast cancer. J Consult Clin Psychol 71:189-200, 2003
Savard J, Simard S, Ivers H, et al: Randomized study on the efficacy of cognitive-behavioral therapy for insomnia secondary to breast cancer, Part II: Immunologic effects. J Clin Oncol 23:6097-6106, 2005
American Sleep Disorders Association: International Classification of Sleep Disorders (ICSD): Diagnostic and Coding Manual. Rochester, MN, American Sleep Disorders Association, 1997
American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders (ed 4). Washington, DC, American Psychiatric Association, 1994
Morin CM: Insomnia: Psychological Assessment and Management. New York, The Guilford Press, 1993
Derogatis LR, Feldstein M, Morrow G, et al: A survey of psychotropic drug prescriptions in an oncology population. Cancer 44:1919-1929, 1979
Stiefel FC, Kornblith AB, Holland JC: Changes in the prescription patterns of psychotropic drugs for cancer patients during a 10-year period. Cancer 65:1048-1053, 1990
Backhaus J, Hohagen F, Voderholzer U, et al: Long-term effectiveness of a short-term cognitive-behavioral group treatment for primary insomnia. Eur Arch Psychiatry Clin Neurosci 251:35-41, 2001
Espie CA, Inglis SJ, Tessier S, et al: The clinical effectiveness of cognitive behaviour therapy for chronic insomnia: Implementation and evaluation of a sleep clinic in general medical practice. Behav Res Ther 39:45-60, 2001
Murtagh DRR, Greenwood KM: Identifying effective psychological treatments for insomnia: A meta-analysis. J Consult Clin Psychol 63:79-89, 1995
Morin CM, Culbert JP, Schwartz SM: Nonpharmacological interventions for insomnia: A meta-analysis of treatment efficacy. Am J Psychiatry 151:1172-1180, 1994
McCrae CS, Lichstein KL: Secondary insomnia: Diagnostic challenges and intervention opportunities. Sleep Med Rev 5:47-61, 2001
McCrae CS, Lichstein KL: Secondary insomnia: A heuristic model and behavioral approaches to assessment, treatment, and prevention. Appl Prev Psychol 10:107-123, 2001
First MB, Spitzer RL, Gibbon M, et al: Structured Clinical Interview for DSM-IV Axis I Disorders: Patient Edition (SCID-I/P, Version 2.0). New York, NY, Biometrics Research Department, New York State Psychiatric Institute, 1996
Coates TJ, Killen JD, George J, et al: Estimating sleep parameters: A multitrait-multimethod analysis. J Consult Clin Psychol 50:345-352, 1982
Rechtschaffen A, Kales A: A Manual of Standardized Terminology: Techniques and Scoring System for Stages in Human Subjects. Bethesda, MD, US Dept of Health, Education, and Welfare, Public Health Service—National Institutes of Neurological Diseases and Blindness, Neurological Information Network, 1968
Savard MH, Savard J, Simard S, et al: Empirical validation of the Insomnia Severity Index in cancer patients. Psychooncology 14:429-441, 2005
Bastien CH, Vallieres A, Morin CM: Validation of the Insomnia Severity Index as an outcome measure for insomnia research. Sleep Med 2:297-307, 2001
Smith S, Trinder J: Detecting insomnia: Comparison of four self-report measures of sleep in a young adult population. J Sleep Res 10:229-235, 2001
Zigmond AS, Snaith RP: The Hospital Anxiety and Depression Scale. Acta Psychiatr Scand 67:361-370, 1983
Roth AJ, Kornblith AB, Batel-Copel L, et al: Rapid screening for psychologic distress in men with prostate carcinoma. Cancer 82:1904-1908, 1998
Savard J, Laberge B, Gauthier JG, et al: Screening clinical depression in HIV-seropositive patients using the Hospital Anxiety and Depression Scale. AIDS Behav 3:167-175, 1999
Savard J, Laberge B, Gauthier JG, et al: Evaluating anxiety and depression in HIV-infected patients. J Pers Assess 71:349-367, 1998
Smets EMA, Garssen B, Bonke B, et al: The multidimensional fatigue inventory (MFI) psychometric qualities of an instrument to assess fatigue. J Psychosom Res 39:315-325, 1995
Fillion L, Gelinas C, Simard S, et al: Validation evidence for the French Canadian adaptation of the Multidimensional Fatigue Inventory as a measure of cancer-related fatigue. Cancer Nurs 26:143-154, 2003
Aaronson NK, Ahmedzai S, Bergman B: The EORTC QLQ-C30: A quality of life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 85:365-376, 1993
Osoba D, Aaronson N, Zee B, et al: Modification of the EORTC QLQ-C30 (version 2.0) based on content validity and reliability testing in large samples of patients with cancer. Qual Life Res 6:103-108, 1997
Tabachnik BG, Fidell LS: Using Multivariate Statistics (ed 4). New York, NY, Harper Collins, 2001
SAS Institute: SAS/STAT User's Guide, Version 8: Volumes 1, 2, and 3. Cary, NC, SAS Institute, 2001
Baron RM, Kenny DA: The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. J Pers Soc Psychol 51:1173-1183, 1986
Keselman HJ, Algina J, Kowalchuk RK: The analysis of repeated measures designs: A review. Br J Math Stat Psych 54:1-20, 2001
Frigon J-Y, Laurencelle L: Analysis of covariance: A proposed algorithm. Educ Psychol Meas 53:1-18, 1993
McClusky HY, Milby JB, Switzer PK, et al: Efficacy of behavioral versus triazolam treatment in persistent sleep-onset insomnia. Am J Psychiatry 148:121-126, 1991
Morin CM, Colecchi C, Stone J, et al: Behavioral and pharmacological therapies for late-life insomnia: A randomized controlled trial. JAMA 281:991-999, 1999
Morin CM, Hauri PJ, Espie CA, et al: Nonpharmacologic treatment of chronic insomnia. Sleep 22:1134-1156, 1999
Jacobs GD, Benson H, Friedman R: Perceived benefits in a behavioral-medicine insomnia program: A clinical report. Am J Med 100:212-216, 1996
Morin CM, Kowatch RA, Barry T, et al: Cognitive-behavior therapy for late-life insomnia. J Consult Clin Psychol 61:137-146, 1993
Trinder J: Subjective insomnia without objective findings: A pseudo diagnostic classification? Psychol Bull 103:87-94, 1988
Morin CM: The nature of insomnia and the need to refine our diagnostic criteria. Psychosom Med 62:483-485, 2000
Wohlgemuth WK, Edinger JD, Fins AI, et al: How many nights are enough? The short-term stability of sleep parameters in elderly insomniacs and normal sleepers. Psychophysiology 36:233-244, 1999
Krystal AD: The changing perspective on chronic insomnia management. J Clin Psychiatry 65:20-25, 2004(Josée Savard, Sébastien S)