Improving mental health and daytime function in adult insomnia patients predict cognitive behavioral therapy for insomnia effectiveness: A case-control study

Abstract

Objective

This study investigated demographic, sleep related symptoms and mental health status as predictors of clinically significant treatment responses to cognitive behavioral therapy in adults who have good adherence for the cognitive behavioral therapy for insomnia (CBT-I) program in primary insomnia.

Methods

A total of 42 adults with primary insomnia disorder were treated with CBT-I at a university hospital from June 2020 to January 2021. Demographic variables were surveyed and sleep-related symptoms were measured using self-reported questionnaires before and after the intervention, comprising a 6-week interval. The treatment responder group was defined as patients with an Insomnia Severity Index change score >7 compared to baseline. Logistic regression and paired t-test examined whether these factors predicted treatment outcomes for CBT-I.

Results

Demographic variables did not predict treatment outcomes. Higher levels of anxiety were associated with a higher likelihood of treatment response (odds ratio [OR] = 1.234; confidence interval [CI]: 1.008–1.511). More severe insomnia at baseline was associated with a greater likelihood of treatment response (OR = 1.450; CI: 1.121–1.875). The lesser the dysfunctional beliefs and attitudes about sleep, the more effective the treatment response (OR = 0.943; CI: 0.904–0.984). Unlike the group of treatment responders, daytime function, depressive mood, and anxiety status did not improve in the group of treatment non-responders after CBT-I intervention.

Conclusions

Patients with severe insomnia and anxiety at baseline should be treated more aggressively with CBT-I. During treatment, patients’ mental health problems and daytime activities should be continuously monitored, in order to help improve these problems which might strengthen the effectiveness of CBT-I.

Highlights

• •Patients with severe insomnia should be treated more aggressively with CBT-I.
• •Mental health and daytime function are important roles in the outcome of CBT-I.
• •Patients with a high anxiety about sleep require more tailored CBT.
• •Patients's daytime activities should be continuously monitored during CBT-I.

1. Introduction

The International Classification of Sleep Disorders, third edition (ICSD-3) diagnostic manual describes chronic insomnia disorder as a difficulty initiating or maintaining sleep or waking up too early with associated daytime consequences, despite adequate opportunity and circumstances for sleep. Sleep difficulties must occur at least three times per week for at least 3 months. [1] A variety of maladaptive cognitions and behaviors play a critical role in the development and maintenance of chronic insomnia. [2] These include performance anxiety and negative expectations regarding sleep, associated concerns about the potential consequences of not sleeping, and unhelpful beliefs and attitudes about sleep. In addition, unhelpful behaviors can have a direct impact on the physiological systems controlling sleep. [3] Based on the pathophysiological mechanism of chronic insomnia, cognitive behavior therapy for insomnia (CBT-I) is the standard and first-line treatment for chronic insomnia, as recommended worldwide by the American Academy of Sleep Medicine, the American College of Physicians [4,5],and the World Sleep Society. [6,7]

Although its effectiveness has been well established, it has been reported that up to 40% of patients do not achieve remission after treatment. [[8], [9], [10]] The reduced effectiveness of CBT-I compared to its efficacy in clinical trials can be explained in part by dropout, poor adherence, and reduced attendance at sessions. [11] For several studies, short sleepers, lower insomnia severity, and greater severity of depression at baseline were related to dropouts in all CBT-I sessions. [12,13] However, in some patients, CBT-I is ineffective despite having well-attended participation and good adherence to treatment. In a few studies on predictors of response to CBT-I, several factors, including short sleep duration, use of hypnotic medications, and the presence of psychopathology, such as depression and anxiety, resulted in less favorable responses to CBT-I. However, there have been somewhat divergent and inconsistent findings, and there were different patient demographics in these studies. [[14], [15], [16]] There is limited research focusing on factors associated with treatment response to CBT-I in patients who have good adherence and attendance for the CBT-I program in primary insomnia.

This study investigated demographic, sleep-related symptoms, and mental health status as predictors of a clinically significant treatment response to CBT-I in adults with primary insomnia who had participated well in all sessions of CBT-I and completed treatment. Through our study, we determined those who are most likely to benefit from CBT-I and which factors are considered to enhance the treatment effect of CBT-I.

2. Material and methods

2.1. Participants

This was a retrospective review of adult patients with primary insomnia who were treated with CBT-I at the outpatient sleep clinic of a single university medical center from June 2020 to August 2021. The study was approved by the CHA medical center Institutional Review Board (IRB approval No.: 2020-06-028). The patients were adults aged >19 years who met the research diagnostic criteria for chronic insomnia according to the ICSD-3. [1] All patients were diagnosed through face-to-face interviews and structured questionnaires with a neurologist (JWS) in the outpatient sleep clinic during their first visit. We included patients who had received a total of four weekly individual face-to-face sessions of CBT-I, with online chat consultation which occurred midday once a week for a total of four weeks with good attendance at the CBT-I program for primary insomnia. ‘Good attendance’ was defined as patients who attended all four face –to face CBT-I sessions, patients for whom sleep schedule and sleep hygiene were checked, and who checked received images and videos related to the treatment program through online chat consultation. Patients were excluded if they had the following: (1) failure to attend all four CBT-I session or to check online consultation during a total of four CBT-I session (2) uncontrolled medical conditions suspected to interfere with sleep or requiring immediate treatment outside of the study; (3) uncontrolled psychiatric condition requiring immediate treatment outside of the study, including current major depressive episodes; and (4) previous diagnosis or evidence of specific sleep disorders, such as obstructive sleep apnea, restless legs syndrome, or circadian rhythm sleep disorders (screened using clinical interview and sleep questionnaires including Sleep-50 [17]).

2.2. Intervention

All patients who received CBT-I attended four weekly individual face-to-face treatment sessions, each lasting approximately 1 h. Patients were also checked for sleep schedule and sleep hygiene and received images and videos related to the treatment program through online chat consultation (Kakao channel, Seoul, Korea) which occurred midday once a week for a total of four weeks. The rationale for the CBT-I condition was based on a manualized multicomponent approach that included several modules introduced at different stages of the treatment process. [18] These components include sleep hygiene education, [19] sleep restriction, [20] stimulus control, [21] cognitive therapy, [22] and progressive relaxation techniques.

The treatment sessions were facilitated by two clinical psychologists (YJS and SYK), who are trained in CBT-I with doctoral degrees (Ph.D.). They supervised a clinical psychologist (SYP) who is certified by the Korean Psychological Association with more than 10 years of CBT practice, who delivered the formalized program curriculum to the participants.

2.3. Treatment predictors

Predictors of treatment outcomes were collected at the baseline assessment and included demographic and clinical characteristics assessed by questionnaires, which were assessed prior to treatment initiation. Treatment response was determined after 4 weeks of treatment, and follow-up self-questionnaires were completed 2 weeks after the final treatment. Self-assessments were administered before treatment and 2 weeks after treatment, which comprised a 6-week interval. Clinical demographics, age, sex, marital or cohabiting status, employment status, and education were assessed by self-reporting. Medical comorbidities were evaluated using the comorbidity questionnaire developed at the Center for Research on Chronic Disorders at the University of Pittsburgh School of Nursing. [23] Information regarding participants’ use of medications known to affect sleeping pills (benzodiazepines, hypnotics, antidepressants, antipsychotics, and anxiolytics) was collected via self-reporting, and we grouped participants into three categories; no use of sleeping pills, use occasionally (less than 3 times a week), and use frequently(more than once every 2 days) or daily. Depressive mood and anxiety were evaluated using the Korean versions of the Patient Health Questionnaire (PHQ-9) [24] and General Anxiety Disorder (GAD-7). [25]

To assess sleep-related symptoms at baseline and status after CBT-I, the Korean version of the Pittsburgh Sleep Quality Index (PSQI), [26] Insomnia Severity Scale (ISI), [27] Dysfunctional Beliefs and Attitudes about Sleep Scale (DBAS-16), [28] and Glasgow Sleep Effort Scale (GSES) [29] were evaluated using a self-questionnaire. Short sleepers were defined as those who slept less than 5 h a day and who were investigated using the questionnaire item of PSQI (e.g., Q4: During the past month, how many hours of sleep did you get at night? (This may differ from the number of hours spent in bed)).

Participants were considered treatment responders if their ISI change score compared with baseline was >7 and as treatment remitters if their absolute ISI score was <8.

2.4. Statistical analysis

For data analysis, we used SPSS (version 26.0; IBM Corp., Armonk, NY, USA). Categorical data were compared using the χ² test and Fisher's exact test. Continuous variables were presented as mean ± standard deviation (SD), and independent t-test was used for comparisons between the two groups. Logistic regression models regressed each individual baseline variable on response criteria. Factors that improved after treatment were investigated using paired t-test in each group with responders and non-responders of CBT-I to determine if there was a difference between the two groups. Differences were considered statistically significant at p < 0.05. A receiver-operating characteristic (ROC) curve analysis with area under the curve (AUC) was used to calculate the prediction accuracy of the ISI and GAD-7 scores for response outcome for CBT-I.

3. Result

A total of 52 patients underwent CBT-I treatment from June 2020 to August 2021 in our hospital, and 42 of 52 completed participation in four weekly individual face-to-face treatment session with online chat consultation. Eight patients did not complete all four CBT-I sessions or did not check in for the online chat consultation. One patient had a history of epilepsy and epilepsy symptoms were aggravated, and one patient attempted suicide during the CBT-I period. Among 42 patients who had good attendance of CBT-I, 25 patients (59.5%) met the criteria for response to CBT-I (Fig. 1).

Fig. 1.

Selection of patients who underwent CBT-I, including inclusion and exclusion criteria as well as subgroup assignment and patient numbers.

At baseline, the mean ± SD age of participants was 52.8 years ± 13.8 and 66.7% were female. There were 73.8% of patients who were married or living with someone, 54.8% graduated 4 years of college, 40.5% were working, and 42.9% had medical comorbidities. There were 16.7% of patients who took sleeping pills daily.

The mean ± SD of ISI was 20.8 ± 4.9, indicating at least moderate insomnia in all patients. Patients had a high level of intrusive thoughts while trying to sleep and a tendency to worry about their problems regarding sleep status, with mild to moderate depressive mood (PHQ-9 mean ± SD:9.7 ± 5.2) and anxiety (GAD-9 mean SD:7.4 ± 5.0). There were no differences in the basal demographics between responders and non-responders of CBT-I. However, patients who were responders of CBT-I had more severe insomnia at baseline than the non-responders (Table 1).

Table 1.

Baseline demographic and clinical characteristics.

Characteristic	No. (%)		P-value
	Responder (N = 25)	Non-responder (N = 17)
Age, mean (SD), y	53.1 (13.8)	50.8 (14.3)	0.617
Sex, female (%)	18 (72.2)	10 (58.8)	0.508
Married or cohabiting (%)	20 (80)	11 (64.7)	0.305
Graduated from 4-y college (%)	13 (52)	13 (58.8)	0.757
Employed (%)	9 (52)	8 (47.1)	0.534
Comorbidity of medical disorders (%)	12 (48)	6 (35.3)	0.530
Taking hypnotics, (%)			0.598
-frequently (more than once every 2 days	5 (20)	2 (11.8)
-occasionally (less than 3 times a week)	12 (48)	9 (52.9)
-not taking hypnotics	8 (32)	6 (35.3)
Mental health
Patient Health Questionnaire-9, mean (SD)	10.32(5.17)	8.71(5.14)	0.197
Generalized Anxiety Disorder-7, mean (SD)	8.56 (5.42)	5.71(3.99)	0.071
Sleep health
Insomnia Severity Index, mean (SD)	22.28 (4.43)	18.76 (5.24)	0.017
Pittsburgh Sleep Quality Index, mean (SD)	14.60 (2.75)	13.82 (3.41)	0.421
Subjective sleep quality, mean (SD)	2.60 (0.70)	2.24 (0.97)	0.166
daytime dysfunction, mean (SD)	1.48 (1.09)	1.65 (1.00)	0.616
sleep disturbance, mean (SD)	1.52 (0.59)	1.59 (0.62)	0.722
Sleep latency, >60 min, n (%)	20 (80)	11 (64.7)	0.305
Sleep duration, <5hr, n (%)	18 (72)	8 (47.1)	0.121
Sleep efficacy, <75%, n (%)	18 (72)	13 (76.5)	1.000
Epworth Sleepiness Scale, mean (SD)	6.16 (4.12)	6.35 (3.74)	0.878
Dysfunctional Beliefs and Attitudes about Sleep scale - 16, mean (SD)	86.40 (32.32)	93.65 (27.20)	0.452
Glasgow Sleep Effort Scale, mean (SD)	8.84 (2.85)	9/06(2.70)	0.805

SD, standard deviation.

Logistic regression models that regressed baseline demographic, clinical, and sleep characteristics on the response criteria are reported in Table 2. A higher level of anxiety was associated with a higher likelihood of treatment response (odds ratio [OR] = 1.234; confidence interval [CI]: 1.008–1.511). In addition, more severe insomnia at baseline, as indicated by higher ISI scores, was associated with a greater likelihood of treatment response (OR = 1.450; CI: 1.121–1.875). The lesser the dysfunctional beliefs and attitudes about sleep, the more effective the treatment response (OR = 0.943; CI: 0.904–0.984). The AUC for ISI and GAD-7 scores were 0.724 (0.581–0.877) and 0.685 (0.518–0.851), respectively (95% confidence interval). This indicates that both ISI and GAD-7 scores are reliable indicators for the predictive values of response outcome in the CBT-I (Fig. 2).

Table 2.

Logistic regression predicting the odds of treatment response after cognitive-behavior therapy for insomnia.

	Odds Ratio (95% CI)
Demographic Characteristics
-Age	0.959 (0.870-1.058)
-Sex, female	0.123 (0.001-27.753)
-Married or cohabiting	0.133 (0.004-4.838)
-Graduated from 4-y college	9.386 (0.001-92797.115)
-Employed	0.162 (0.024-1.112)
Clinical characteristics
-Comorbidity of medical disorders	2.688 (0.222-32.601)
-Taking hypnotics (occasionally)	0.108 (0.002-5.882)
-Taking hypnotics (frequently)	0.336 (0.004-26.573)
-Anxiety	1.234 (1.008–1.511)∗
-Depressive mood	0.794 (0.546-1.156)
Sleep characteristics
-Insomnia severity	1.450 (1.121–1.875)∗
-Sleep quality	0.620 (0.259-1.488)
-Dysfunctional Beliefs and Attitudes about Sleep	0.943 (0.904–0.984)∗
-Sleep effort	0.742 (0.479-1.151)
-Subjective short sleeper	5.483 (0.603-49.889)

∗p ≤ 0.05; CI, confidence interval.

Fig. 2.

ROC analysis of ISI and GAD-7 scores for predictive values of response outcomes in the CBT-I.

After treatment with CBT-I, all sleep-related symptoms, including quality of sleep and dysfunctional beliefs and attitudes about sleep, significantly improved in both groups. However, in the non-responder group, although sleep-related symptoms improved, depression and anxiety did not improve, and daytime dysfunction, which is a sub-item of the PSQI, did not improve (Table 3 and Fig. 3).

Table 3.

Comparison of changes in questionnaire scores before and after CBT-I treatment between groups with responders and non-responders.

Mean (SD)	Responder			Non- responder
	Pre-treatment	Post-treatment	P-value	Pre-treatment	Post-treatment	P-value
PSQI	14.60 (2.75)	10.24 (3.49)	<0.001	12.64 (3.33)	10.0 (1.32)	0.009
Subjective sleep quality	2.58 (0.72)	1.54 (0.779)	<0.001	2.0 (1.09)	1.27 (0.91)	0.004
Sleep disturbance	1.50 (0.59)	1.17 (0.38)	<0.001	1.55 (0.522)	1.09 (0.30)	0.016
Daytime dysfunction	1.50 (1.10)	0.96 (0.86)	<0.045	1.45 (0.93)	1.27 (0.65)	0.441
ESS	5.91 (4.03)	4.04 (3.25)	0.065	6.17(3.69)	6.00(2.89)	0.772
DBAS-16	86.75 (35.93)	47.50 (35.93)	<0.001	95.50 (31.10)	62.25 (34.49)	0.001
Glasgow Sleep Effort Scale	8.96 (2.76)	5.57 (3.98)	<0.001	8.75 (2.96)	5.58 (3.66)	<0.001
PHQ-9	10.32 (5.17)	4.36 (3.26)	<0.001	9.58 (5.57)	7.75 (3.96)	0.209
GAD-7	8.56 (5.42)	4.76 (4.44)	0.001	5.58 (3.70)	5.50 (4.34)	0.950

CBT-I, cognitive behavioral therapy for insomnia; SD, standard deviation; PSQI, Pittsburgh Sleep Quality Index; ESS, Epworth Sleepiness Scale; DBAS-16, Dysfunctional Beliefs and Attitudes about Sleep scale – 16; GSES, Glasgow Sleep Effort Scale; PHQ-9, Patient Health Questionnaire-9; GAD-7, Generalized Anxiety Disorder-7.

Fig. 3.

Changes of questionnaire scores before and after cognitive behavioral therapy for insomnia treatment in groups with responders and non-responders.

4. Discussion

In our study, approximately 40% of patients with primary insomnia who received CBT-I did not achieve remission after treatment. At baseline demographics, we determined no significant influence of age, sex, hypnotics use including patterns of hypnotics usage, marriage, education, and medical comorbidity on outcomes of CBT-I. Rather, the severity of insomnia and psychological distress, such as anxiety at baseline, were predictors of treatment response to CBT-I. Moreover, patients did not feel that their insomnia improved unless psychological distress, such as depression, anxiety, and daytime function, improved. Women and older adults tend to express a higher risk for chronic insomnia in the general population. [30,31] A strength of the current study was that to maximize generalizability of the findings to adults with chronic insomnia, and the majority of participants in our study were older adults (participants >50 years old, 65.9%) and were predominantly women (66.7%). Additionally, we did not exclude participants with stably treated co-occurring medical or psychiatric conditions or patients who had taken hypnotics. The presence of such comorbidities and hypnotic use including patterns of hypnotics usage did not predict CBT-I treatment response in this study.

Baseline demographics were not significant predictors of treatment response to CBT-I in our study, and this finding replicates previous data on predictors. [16,32] Hypnotic using patients responded equally well to CBT-I in the previous study [16], as same as our study. Also, patterns of hypnotics usage did not affect outcomes of CBT-I response. However, our study had small samples, and there was no study to investigate whether intermittent hypnotics use during CBT-I has a good or poor outcome of treatment response. Therefore further studies for this issue will be necessary. In a previous study, older age was associated with better attendance and adherence to consistent sleep and wake times, and they were more conscientious. [33] We hypothesized predictors of good adherence and attendance of the CBT-I program would be different from those for good treatment outcome of CBT-I. A recent study, including our results, showed that age did not affect treatment response to CBT-I. [11]

Mood symptom-related insomnia has different roles as predictors of adherence and treatment response in CBT-I. For example, the more severe the anxiety and depression, the less likely they were to be able to participate effectively in treatment until the end of the CBT-I program and drop out. [12,13] However, the more severe the insomnia symptoms and sleep-related anxiety, the better the effect of CBT-I treatment response. Therefore, severe symptoms of insomnia and psychological distress are predictors of good treatment response to CBT-I. This finding is consistent with the results of several previous studies. [32,34] Good adherence and attendance in programs decrease as anxiety and depression become more severe; however, the therapeutic effect may be rather effective when these patients finish the CBT-I program. Therefore, when dealing with patients with severe psychological distress related to sleep in CBT-I, it is necessary for the therapist to carefully consider them so that they can better follow the treatment. When we treated CBT-I at our clinic, in addition to the treatment program once a week, we checked whether patients were keeping their promises through an online chat consultation (Kakao channel, Seoul, Korea) which occurred midday once a week, and communicated with patients by sending the contents of the program collected the previous week to a file. Therefore, in our study, even patients with high anxiety were able to follow up until the end, and it is thought that the therapeutic effect was good.

In our study, the less severe the dysfunctional beliefs and attitudes about sleep, the better the treatment response to CBT-I. The results of this study were inconsistent with those of previous studies. Previous research has shown that dysfunctional thinking is a positive indication for CBT-I. [16,35] However, Jansson-Fröjmark and Linton concluded from their study of CBT-I that sleep-related beliefs play a less predictive role than previously envisioned. [36] Another recent study supported that the DBAS, Sleep Problems Acceptance Questionnaire, and Sleep-Related Behaviors Questionnaire would not be unique predictors of outcome for CBT-I. [37]

In one study, the relationship between the treatment response of CBT-I and dysfunctional beliefs of sleep was analyzed by dividing it into subgroups using DBAS questionnaires. Theme or subscale scores were then calculated by taking the mean scores on the items comprising the following DBAS-16 subscales: (1) expectations for sleep (two items), (2) worry/helplessness about sleep (six items), (3) consequences of insomnia (five items), and (4) medication (three items). [38] It appears that insomnia sufferers fitting the ‘‘worried and symptom-focused’’ profile are good responders to the form of CBT delivered in this study. However, the “worried and medication-biased’’ group remained above the norm for insomnia symptoms after CBT-I. The belief ‘‘I'd better take sleeping pills'’ may need other types of intervention rather than the psychoeducation component included in the current form of CBT.

Considering this study, primary insomnia is characterized by heterogeneous patterns of unhelpful beliefs. Therefore, patterns of unhelpful beliefs about sleep may predict different responses to CBT-I. Although our study did not divide the DBAS into subgroups, future studies will focus on this and increase the number of participants to determine what dysfunctional thoughts about sleep can actually predict the therapeutic effect of CBT-I.

The strength of our study is that unlike previous studies, we investigated which factors did not improve in the groups with CBT-I responders and non-responders. Interesting results were shown in this analysis, and the group with non-responders showed improvement in all symptoms related to sleep, including sleep quality, sleep disturbance, dysfunctional beliefs of sleep, and sleep effort, similar to the group with CBT-I responders. However, the group of non-responders showed no improvement in depression, anxiety, and daytime function. If distress and mood in everyday life do not improve, the patients themselves evaluate that the symptoms of insomnia have not improved. In one study, insomnia patients with stress and depression symptoms maintained the highest percentage of clinical depression at the end-of-treatment and insomnia at long-term follow-up (7.8 ± 1.6 years after the end-of-treatment) in comparison with those who did not have mood and distress symptoms. [39] In another study, they assessed the value of the subtypes for predicting responses to online CBT-I provided for a subsample of 42 insomnia patients compared with 26 patients in a waitlist control group. In subjects classified as “moderately distressed but with intact responses to pleasurable emotions”, CBT-I led to a large decrease in ISI score and SL. On the other hand, in subjects classified as “slightly distressed with high reactivity to their environment and life events”, the ISI score decreased, but SL showed no improvement. [40] Additionally, in our study, as shown in Fig. 3, GAD-7, a measure of anxiety, seemed to be a factor that influenced response of CBT-I more than PHQ-9, a measure of depression. A potential mechanism underlying anxiety-insomnia association is the role of worry, which is a phenomenological feature of generalized anxiety disorder and a critical element in the cognitive model of insomnia. Specifically, inappropriate worry about sleep and sleep-related consequences may lead to arousal, which perpetuates insomnia. [41] Given this connection between worry and sleep, for patients who continue to have poor sleep that does not improve and persist during the treatment of CBT-I, anxiety might increase during process of emphasizing sleep restriction and repetitive modifications of dysfunctional beliefs in CBT-I. As a result, increased anxiety itself could hinder treatment effects of CBT-I.

Considering these results, mental health status and distress symptoms should be considered as factors that play an important role in the outcome of CBT-I. Therefore, patients with poor mental health status, distress, and higher worry and beliefs about sleep require more tailored cognitive behavior therapy. Therefore, when CBT-I is practiced in patients, it is important to focus on mental health and daytime function. Scheduling activities and rewards in actual psychotherapy for patients with depression could be addressed in behavioral therapy for patients with insomnia.

During the COVID-19 pandemic, not only daytime stress, including reduced social interaction, being forced to stay at home, anxiety, and depression, but also disrupted sleep was increased. [42] A study suggested that emotional regulation strategies such as rumination, cognitive reappraisal, and suppression play a mediating role in sleep and psychological stress during the COVID-19 pandemic. [43] Mindfulness intervention with CBT-I may be helpful for people with poor emotional regulation. A previous pilot study showed a significant improvement in sleep-related quality of life in non-responders to CBT-I with chronic primary insomnia immediately after six weekly outpatient sessions of acceptance and commitment therapy, including mindfulness, and at the 3-month follow-up. [44]

Several studies on the response of CBT-I to short sleepers have been conducted; however, objective or subjective sleep duration has been found to be an inconsistent predictor of CBT-I. [11] In our study, there was no difference in the proportion of short sleepers between groups with responders and non-responders of CBT-I. However, our study had some limitations. First, we did not analyze objective sleep time parameters, such as a sleep diary or actigraphy. Since the sleep duration was obtained from the PSQI, there are limitations in interpreting sleep duration as a predictor of CBT-I outcome with our results. Second, the sample size was small; therefore, larger studies are necessary to confirm our results. The analysis of sleep diary-derived data could have added strength to our conclusions. Furthermore, diary data would shed light on actual therapy compliance and could add an essential variable to the prediction model. Nevertheless, this is a meaningful study that searched for therapeutic responses with data from patients who completed treatment by providing ‘online chat feedback’ during treatment.

5. Conclusion

As in previous studies, patients with severe insomnia and high anxiety level may be more responsive to CBT-I independent of demographic characteristics. During treatment, patients’ mental health problems and daytime activities should be continuously monitored, in order to improve these problems which might strengthen the effectiveness of CBT-I.

Funding

This research received no external funding.

Institutional Review Board statement

The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Ethics Committee of the CHA medical University (ID: 2020-06-028).

CRediT authorship contribution statement

Jung-Won Shin: contributed to study conceptualization, recruited the participants, performed data analysis and prepared the manuscript, supervised the study, Conceptualization, Formal analysis, Data curation, Supervision. Seonyeop Kim: were involved in the management of participants, performed data analysis and prepared the manuscript, Supervision, Formal analysis, Data curation. Bomi Park: were involved in the management of participants, All authors have read and agreed to the published version of the manuscript. Yoon Jung Shin: were involved in the management of participants. Sunyoung Park: were involved in the management of participants, supervised the study, Supervision.

Abbreviations

CBT-I

cognitive behavior therapy for insomnia

ISI

Insomnia Severity Index

PSQI

Pittsburgh Sleep Quality Index

DABS

Dysfunctional Beliefs and Attitudes about Sleep scale

GSES

Glasgow Sleep Effort Scale

PHQ

Patient Health Questionnaire

GAD

Generalized Anxiety Disorder