Hostname: page-component-76fb5796d-r6qrq Total loading time: 0 Render date: 2024-04-26T22:54:33.473Z Has data issue: false hasContentIssue false

Improved resilience following digital cognitive behavioral therapy for insomnia protects against insomnia and depression one year later

Published online by Cambridge University Press:  08 March 2022

Philip Cheng*
Affiliation:
Thomas Roth Sleep Disorders and Research Center, Henry Ford Health System, 39450 W 12 Mile Road, Novi, MI 48197, USA
David A. Kalmbach
Affiliation:
Thomas Roth Sleep Disorders and Research Center, Henry Ford Health System, 39450 W 12 Mile Road, Novi, MI 48197, USA
Hsing-Fang Hsieh
Affiliation:
Department of Health Behavior and Health Education, School of Public Health, University of Michigan, 1415 Washington Heights, Ann Arbor, MI 48109 USA
Andrea Cuamatzi Castelan
Affiliation:
Thomas Roth Sleep Disorders and Research Center, Henry Ford Health System, 39450 W 12 Mile Road, Novi, MI 48197, USA
Chaewon Sagong
Affiliation:
Thomas Roth Sleep Disorders and Research Center, Henry Ford Health System, 39450 W 12 Mile Road, Novi, MI 48197, USA
Christopher L. Drake
Affiliation:
Thomas Roth Sleep Disorders and Research Center, Henry Ford Health System, 39450 W 12 Mile Road, Novi, MI 48197, USA
*
Author for correspondence: Philip Cheng, E-mail: pcheng1@hfhs.org
Rights & Permissions [Opens in a new window]

Abstract

Background

While the negative consequences of insomnia are well-documented, a strengths-based understanding of how sleep can increase health promotion is still emerging and much-needed. Correlational evidence has connected sleep and insomnia to resilience; however, this relationship has not yet been experimentally tested. This study examined resilience as a mediator of treatment outcomes in a randomized clinical trial with insomnia patients.

Methods

Participants were randomized to either digital cognitive behavioral therapy for insomnia (dCBT-I; n = 358) or sleep education control (n = 300), and assessed at pre-treatment, post-treatment, and 1-year follow-up. A structural equation modeling framework was utilized to test resilience as a mediator of insomnia and depression. Risk for insomnia and depression was also tested in the model, operationalized as a latent factor with sleep reactivity, stress, and rumination as indicators (aligned with the 3-P model). Sensitivity analyses tested the impact of change in resilience on the insomnia relapse and incident depression at 1-year follow-up.

Results

dCBT-I resulted in greater improvements in resilience compared to the sleep education control. Furthermore, improved resilience following dCBT-I lowered latent risk, which was further associated with reduced insomnia and depression at 1-year follow-up. Sensitivity analyses indicated that each point improvement in resilience following treatment reduced the odds of insomnia relapse and incident depression 1 year later by 76% and 65%, respectively.

Conclusions

Improved resilience is likely a contributing mechanism to treatment gains following insomnia therapy, which may then reduce longer-term risk for insomnia relapse and depression.

Type
Original Article
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
Copyright © The Author(s), 2022. Published by Cambridge University Press

Introduction

As the most prevalent and burdensome sleep disorder (Ohayon, Reference Ohayon2002; Roth, Reference Roth2007; Roth & Roehrs, Reference Roth and Roehrs2003), insomnia is a significant public health concern. Chronic insomnia impairs daytime functioning, wreaks havoc on multiple domains of health, and increases the risk of medical morbidities more than fivefold compared to healthy sleepers (Kalmbach, Pillai, Arnedt, & Drake, Reference Kalmbach, Pillai, Arnedt and Drake2016b; Taylor et al., Reference Taylor, Mallory, Lichstein, Durrence, Riedel and Bush2007). Left untreated, insomnia is also a robust risk factor for mental illness (Pigeon, Bishop, & Krueger, Reference Pigeon, Bishop and Krueger2017), including depression (OR 3.95), substance use disorders (OR 7.18) (Breslau, Roth, Rosenthal, & Andreski, Reference Breslau, Roth, Rosenthal and Andreski1996), and even suicide controlling for depression (OR 1.30) (Bernert, Turvey, Conwell, & Joiner, Reference Bernert, Turvey, Conwell and Joiner2014).

Though the destruction of insomnia has been well-documented, the contrast of a strengths-based approach to understanding the role of sleep in health promotion is still emergent and much-needed (Buysse, Reference Buysse2014). Nascent evidence suggests that sleep promotes protective physiological functions such as clearing of harmful neuronal waste (e.g. β-amyloid) (Winer et al., Reference Winer, Mander, Kumar, Reed, Baker, Jagust and Walker2020) or myelination of neurons (Bellesi, Bushey, Chini, Tononi, & Cirelli, Reference Bellesi, Bushey, Chini, Tononi and Cirelli2016; Bellesi et al., Reference Bellesi, Haswell, de Vivo, Marshall, Roseboom, Tononi and Cirelli2018). In contrast to a deficit model, a strengths-based approach examines whether promotion of sleep health may interrupt and prevent a worsening health trajectory, or perhaps even alter the trajectory toward an ‘upward spiral’ that eventually returns to or surpasses premorbid levels of health and functioning.

In a strengths-based approach, resilience is of particular interest because it is among the critical components that determine health trajectories following stressors (Bonanno, Reference Bonanno2005). Though there are various definitions of psychological resilience, a core component is the ability to respond and recover from stressors (i.e. the root of ‘resilience’ is ‘resile’, which means ‘to bounce or spring back’) (Carver, Reference Carver1998; Tusaie & Dyer, Reference Tusaie and Dyer2004). Resilience has been linked to maintaining health, wellbeing, and even thriving despite facing stressors. Resilience is also of interest as a therapeutic target because it represents an opportunity for prevention and early intervention. Under the diathesis-stress model of psychopathology, stress in combination with predispositional risk factors precipitates psychopathology; however, individuals with higher levels of resilience may be more likely to respond adaptively and recover from stress. This then protects against adverse outcomes via reduced intensity and duration of symptoms, or even avoiding illness altogether.

There is growing evidence that sleep and resilience are interconnected. First, those who report greater psychological resilience also exhibit more robust sleep (e.g. greater sleep efficiency, fewer and shorter awakenings after sleep onset, less light sleep, and more deep sleep) (Brand et al., Reference Brand, Gerber, Kalak, Kirov, Lemola, Clough and Holsboer-Trachsler2014a, Reference Brand, Gerber, Kalak, Kirov, Lemola, Clough and Holsboer-Trachsler2014b, Reference Brand, Kalak, Gerber, Clough, Lemola, Pühse and Holsboer-Trachsler2016). Similarly, those with poor sleep also report less resilience. Not only do good sleepers report greater resilience than those with insomnia, high levels of resilience protected individuals against the effects of stress on sleep quality (Palagini et al., Reference Palagini, Moretto, Novi, Masci, Caruso, Drake and Riemann2018). This is consistent with patient reports of being more easily frustrated during episodes of insomnia, and with evidence of poor cognitive and emotional coping (Morin, Rodrigue, & Ivers, Reference Morin, Rodrigue and Ivers2003; Pillai, Roth, Mullins, & Drake, Reference Pillai, Roth, Mullins and Drake2014; Voss, Kolling, & Heidenreich, Reference Voss, Kolling and Heidenreich2006). Greater resilience may lead to appraisal of daily life events as less stressful (Morin et al., Reference Morin, Rodrigue and Ivers2003) and decreased occurrence or prolongation of stressful life events (Hall Brown, Akeeb, & Mellman, Reference Hall Brown, Akeeb and Mellman2015; Healey et al., Reference Healey, Kales, Monroe, Bixler, Chamberlin and Soldatos1981; Luo, Zhang, & Pan, Reference Luo, Zhang and Pan2013).

Under the stress-diathesis model, a lack of protective resilience allows stress – and maladaptive appraisals of stress – to serve as a precipitant of psychopathology. This includes the ‘3-P’ model of insomnia (Spielman, Caruso, & Glovinsky, Reference Spielman, Caruso and Glovinsky1987), which posits that insomnia is triggered by a combination of predispositional [e.g. sleep reactivity (Drake, Friedman, Wright, & Roth, Reference Drake, Friedman, Wright and Roth2011), familial risk (Drake, Cheng, Almeida, & Roth, Reference Drake, Cheng, Almeida and Roth2017)] and precipitating factors (e.g. life stressors), which are then maintained by some perpetuating factors [e.g. rumination and poor coping (Pillai et al., Reference Pillai, Roth, Mullins and Drake2014)]. Once elicited, insomnia further potentiates risk for other forms of psychopathologies, including depression (Baglioni et al., Reference Baglioni, Battagliese, Feige, Spiegelhalder, Nissen, Voderholzer and Riemann2011; Li, Wu, Gan, Qu, & Lu, Reference Li, Wu, Gan, Qu and Lu2016), anxiety (Dolsen, Asarnow, & Harvey, Reference Dolsen, Asarnow and Harvey2014), and other psychiatric and medical conditions (Dolsen et al., Reference Dolsen, Asarnow and Harvey2014; Ford & Kamerow, Reference Ford and Kamerow1989; Taylor, Lichstein, & Durrence, Reference Taylor, Lichstein and Durrence2003; Taylor et al., Reference Taylor, Mallory, Lichstein, Durrence, Riedel and Bush2007). Conversely, bolstering resilience may confer protection against stressors and interrupt the development of psychopathology; however, extant studies examining the relationship between sleep and resilience have largely been observational and correlational, thus limiting causal and/or directional inferences.

There is emerging evidence to suggest that resilience can be bolstered by improving sleep, which then confers protection against additional psychopathologies. For example, several studies have shown that cognitive behavioral therapy for insomnia (CBT-I) – the recommended first-line treatment for insomnia (Qaseem, Kansagara, Forciea, Cooke, & Denberg, Reference Qaseem, Kansagara, Forciea, Cooke and Denberg2016; Riemann et al., Reference Riemann, Baglioni, Bassetti, Bjorvatn, Dolenc Groselj, Ellis and Spiegelhalder2017) – has positive effects for comorbid depression despite the fact that CBT-I does not directly target depression and typically has far fewer sessions compared to CBT for depression. More recent studies have shown that treatment of insomnia also prevents the incidence of depression (Batterham et al., Reference Batterham, Christensen, Mackinnon, Gosling, Thorndike, Ritterband and Griffiths2017; Cheng et al., Reference Cheng, Kalmbach, Tallent, Joseph, Espie and Drake2019). Furthermore, we recently demonstrated that individuals with insomnia who received digital CBT-I prior to the COVID-19 pandemic exhibited less stress and depression during the pandemic relative to a control group despite comparable disruptions due to the pandemic (Cheng, Casement, Kalmbach, Castelan, & Drake, Reference Cheng, Casement, Kalmbach, Castelan and Drake2020a). These findings provided further support that improving insomnia may bolster resilience against future stress. However, the direct impact of CBT-I on resilience has not been adequately tested, nor has resilience been tested as a mechanism for positive outcomes following CBT-I. Without this, we are not able to establish resilience as a modifiable factor that can be targeted via sleep intervention to protect against a trajectory of cascading psychopathology.

To expand our strengths-based understanding of sleep health, this study examined resilience in adults with insomnia before and after receiving digital CBT-I (dCBT-I). The central hypothesis was that those who received dCBT-I would report greater resilience compared to those in the control group, and that improved resilience would confer protection from insomnia and depression at 1-year follow-up by reducing risk factors associated with these conditions. We first hypothesized that those in the dCBT-I group would report greater improvements in psychological resilience compared to those in the sleep education control. We then improved resilience from dCBT-I would further mitigate the risk factors of insomnia and depression, which would in turn confer protection from insomnia and depression at 1-year follow-up.

Methods

Data for this study were obtained from participants recruited from six hospitals, 38 medical centers, and subscribers of a major health insurance company in southeastern Michigan. Recruitment occurred between 2016 and 2017, and utilized Internet-based methods (see Cheng et al., Reference Cheng, Luik, Fellman-Couture, Peterson, Joseph, Tallent and Roth2018, Reference Cheng, Kalmbach, Tallent, Joseph, Espie and Drake2019 for more details). Eligible participants met DSM-5 (American Psychiatric Association, 2013) diagnostic criteria for chronic insomnia disorder. Exclusion criteria included diagnosed sleep disorders other than insomnia (e.g. restless legs, narcolepsy) or untreated obstructive sleep apnea, and diagnosed bipolar disorder or seizure disorder. Because the SPREAD trial (NCT02988375) included a depression prevention aim, individuals with high depression chronicity (self-reported daily or near daily depressed mood and anhedonia) were excluded. Those who reported suicidality during screening were further assessed using the Columbia-Suicide Severity Rating Scale (C-SSRS) via telephone within 24 h by research staff certified in conducting the C-SSRS and referred to psychiatric or emergency services when appropriate.

Study design

This study utilized a randomized controlled design with simple randomization into two parallel arms: dCBT-I, or a control condition (online sleep education). Randomization was computerized and automated immediately after participants met eligibility criteria. A total of 1385 individuals with insomnia disorder were enrolled and randomized into either the dCBT-I or online sleep education conditions (see Fig. 1 for enrollment flow chart). The research staff was blinded to treatment allocation. Participants were randomized at a 2:1 ratio for the dCBT-I condition due to a higher anticipated attrition rate for the active compared to the sleep education condition, as has been previously demonstrated in nearly all Internet-based interventions (Christensen, Griffiths, & Farrer, Reference Christensen, Griffiths and Farrer2009). The final sample included 358 participants in the dCBT-I treatment and 300 in the online sleep education condition (see Cheng, Kalmbach, Castelan, Murugan, & Drake, Reference Cheng, Kalmbach, Castelan, Murugan and Drake2020b; Cheng et al., Reference Cheng, Luik, Fellman-Couture, Peterson, Joseph, Tallent and Roth2018 for additional details regarding study design and recruitment). All procedures were approved by the Institutional Review Board. Informed consent was also given by all participants immediately following study eligibility, before any study procedures were implemented.

Fig. 1. Enrollment flow chart.

dCBT-I condition

Individuals randomized to the dCBT-I condition completed the Sleepio program via the Internet (www.sleepio.com, Big Health Ltd). Sleepio was selected among several programs for dCBT-I because it is evidence-based, fully automated, is standardized, and has been tested in almost 7000 participants via multiple RCTs (Barnes, Miller, & Bostock, Reference Barnes, Miller and Bostock2017; Bostock, Luik, & Espie, Reference Bostock, Luik and Espie2016; Cheng et al., Reference Cheng, Luik, Fellman-Couture, Peterson, Joseph, Tallent and Roth2018; Espie et al., Reference Espie, Emsley, Kyle, Gordon, Drake, Siriwardena and Luik2018, Reference Espie, Kyle, Williams, Ong, Douglas, Hames and Brown2012; Freeman et al., Reference Freeman, Sheaves, Goodwin, Yu, Nickless, Harrison and Espie2017; McGrath et al., Reference McGrath, Espie, Power, Murphy, Newell, Kelly and O'Donnell2017; Pillai et al., Reference Pillai, Anderson, Cheng, Bazan, Bostock, Espie and Drake2015). Participants were provided each of the six core sessions of dCBT-I on a weekly basis, which were directed by an animated ‘virtual therapist’ who reviews and guides progress with the participant. Participants had access to the program for up to 12 weeks. The mean attendance rate was 4.1 sessions, and 43% completed all six sessions. The sessions predominantly targeted behavioral and cognitive components of insomnia (e.g. sleep restriction, stimulus control, cognitive restructuring, paradoxical intention), and included relaxation strategies (e.g. progressive muscle relaxation and autogenic training) and sleep hygiene in the first few sessions. The final session focused on relapse prevention.

Online sleep education

Individuals randomized to the online sleep education condition received six weekly e-mails based on the NIH guide to healthy sleep (National Heart Lung and Blood Institute, 2011) containing information on the following topics: the basics of endogenous sleep regulation; the impact on sleep of health problems such as obesity, diabetes, and hypertension; the effects of sleep disruptive substances, such as caffeine, nicotine, alcohol; and tips on creating a sleep-conducive bedroom environment. Psychoeducation and sleep hygiene were selected because they are common in clinical practice, particularly in primary care (Buysse et al., Reference Buysse, Reynolds, Kupfer, Thorpy, Bixler, Kales and Roth1997), and are commonly used as an attention control in clinical trials. Importantly, these are not considered effective standalone treatments for insomnia (Morgenthaler et al., Reference Morgenthaler, Kramer, Alessi, Friedman, Boehlecke and Brown2006).

Measures of interest

Resilience was measured using the Brief Resilience Scale (BRS) (Smith et al., Reference Smith, Dalen, Wiggins, Tooley, Christopher and Bernard2008), which measures one's ability to recover from stress. The BRS contains six items with responses on a five-point Likert scale. The items describe recovery in response to stress (e.g. ‘It does not take me long to recover from a stressful event.’), and responses range from Strong Disagree (1) to Strongly Agree (5). The BRS is scored using the mean of the six items with higher scores indicating greater psychological resilience. The BRS shows good test-retest reliability and construct validity (Smith et al., Reference Smith, Dalen, Wiggins, Tooley, Christopher and Bernard2008). The BRS also showed strong internal reliability in this study (Cronbach's α = 0.89). To support causal inferences in the mediation model via temporal precedence, psychological resilience as a mediator was operationalized as a change score from pre-treatment to post-treatment.

Latent risk for insomnia and depression

In accordance with our interest in resilience as a transdiagnostic construct, we approach the latent risk factor also as a transdiagnostic construct. We were particularly interested in the protection that dCBT-I can confer beyond its well-established effects on symptom reduction. As such, we selected indicators risk based on a stress-diathesis model (in alignment with the 3-P model of insomnia) that also had prior evidence supporting a relationship with both insomnia and depression. Specifically, we examined how changes in resilience following dCBT-I impacted risk as measured by the confluence of transdiagnostic predisposing, precipitating, and perpetuating factors impacting both insomnia and depression. Under this model, all three factors (referred to as the ‘3-Ps’ in the 3-P model) are requisite and compound to increase the incidence of psychopathologies. Sleep reactivity (i.e. vulnerability to sleep disturbance) was selected as a predisposing factor as it has been demonstrated as a heritable predictor of both insomnia and depression (Drake et al., Reference Drake, Friedman, Wright and Roth2011; Drake, Pillai, & Roth, Reference Drake, Pillai and Roth2014; Harvey, Gehrman, & Espie, Reference Harvey, Gehrman and Espie2014; Kalmbach, Anderson, & Drake, Reference Kalmbach, Anderson and Drake2018; Kalmbach, Pillai, Cheng, Arnedt, & Drake, Reference Kalmbach, Pillai, Cheng, Arnedt and Drake2015; Palagini et al., Reference Palagini, Cipollone, Masci, Novi, Caruso, Kalmbach and Drake2019). Stress was selected as a precipitating factor. Finally, a transdiagnostic measure of rumination was selected as a perpetuating factor (Ballesio, Ottaviani, & Lombardo, Reference Ballesio, Ottaviani and Lombardo2019; Carney, Harris, Moss, & Edinger, Reference Carney, Harris, Moss and Edinger2010; Malmberg & Larsen, Reference Malmberg and Larsen2015; Nolen-Hoeksema, Reference Nolen-Hoeksema1991). These three variables were treated as indicators of a latent transdiagnostic risk factor that was then used as a predictor of insomnia and depression at 1-year follow-up.

Sleep reactivity describes a sleep system that is particularly vulnerable to perturbation. Sleep reactivity has been determined as a phenotype that shows stability over time (Jarrin, Chen, Ivers, Drake, & Morin, Reference Jarrin, Chen, Ivers, Drake and Morin2016) and is 30–40% heritable, suggesting a substantial genetic component (Drake et al., Reference Drake, Friedman, Wright and Roth2011). Sleep reactivity is measured using the Ford Insomnia Response to Stress Test (FIRST) (Drake, Jefferson, Roehrs, & Roth, Reference Drake, Jefferson, Roehrs and Roth2006). The FIRST is a self-report measure of sleep reactivity that asks respondents to rate the likelihood (not, somewhat, moderately, and very likely) that they would experience sleep difficulties in reaction to nine hypothetical stressful situations (e.g. ‘after a stressful experience during the day’, ‘before an important meeting the next day’). Higher FIRST scores indicate a more highly sensitive sleep system. Participants with FIRST scores of 15 or below were characterized as having low sleep reactivity, and those who scored 16 or higher were characterized as having high sleep reactivity. The FIRST showed strong internal reliability (Cronbach's α = 0.88).

Stress was measured using a validated single-item instrument (Elo, Leppänen, & Jahkola, Reference Elo, Leppänen and Jahkola2003; Salminen, Kouvonen, Koskinen, Joensuu, & Väänänen, Reference Salminen, Kouvonen, Koskinen, Joensuu and Väänänen2014). The prompt for this instrument was ‘Stress means a situation in which a person feels tense, restless, nervous or anxious or is unable to sleep at night because his/her mind is troubled all the time. Do you feel this kind of stress these days?’ Response on this instrument was on a five-point Likert scale ranging from never (0) to always (4), with higher scores indicating more stress.

Rumination was measured using the Perseverative Thinking Questionnaire (PTQ) (Ehring et al., Reference Ehring, Zetsche, Weidacker, Wahl, Schönfeld and Ehlers2011). The PTQ was selected as a transdiagnostic measure of rumination. The PTQ assesses how often individuals engage in repetitive negative thinking in response to negative experiences or problems using 15 items scored on a five-point Likert scale (0 = Never, 4 = Almost always). The PTQ has strong reliability and convergent validity with other validated rumination scales (e.g. Response Style Questionnaire, Rumination Scale), and with the measures of depression and anxiety (e.g. Beck Depression Inventory, Inventory of Depressive Symptomatology, State Trait Anxiety Inventory, Penn State Worry Questionnaire). As expected, the PTQ showed strong internal reliability in this study (Cronbach's α = 0.96).

Insomnia severity was measured using the Insomnia Severity Index (ISI) (Morin, Belleville, Bélanger, & Ivers, Reference Morin, Belleville, Bélanger and Ivers2011; Thorndike et al., Reference Thorndike, Ritterband, Saylor, Magee, Gonder-Frederick and Morin2011), with higher scores indicating increased insomnia severity (range 0–28). A score of 15 or higher is considered clinically significant insomnia. Items on the ISI in this study showed strong internal reliability (Cronbach's α = 0.89). Depression was measured using the 16 item self-report Quick Inventory of Depressive Symptomatology (QIDS-SR16). This measure also showed strong internal reliability in this study (Cronbach's α = 0.81). The QIDS-SR16 assesses the severity of the nine diagnostic symptom criteria used in the Diagnostic and Statistical Manual of Mental Disorders, and shows strong consistency with a diagnosis of major depressive disorder via a Structured Interview for the DSM (Lamoureux et al., Reference Lamoureux, Linardatos, Fresco, Bartko, Logue and Milo2010; Surís, Holder, Holliday, & Clem, Reference Surís, Holder, Holliday and Clem2016; Yeung et al., Reference Yeung, Feldman, Pedrelli, Hails, Fava, Reyes and Mundt2012). To examine changes in non-sleep symptoms of depression, analyses only utilized non-sleep items from the QIDS (items 1–4 were dropped) unless otherwise specified. Clinically significant depression was determined via a psychometrically derived threshold of 11 or higher on the total score across all items (corresponding to moderate severity or higher). Clinically significant depression was determined using a psychometrically derived threshold of 11 or higher (corresponding to moderate severity or higher). This cutoff has 82.4% sensitivity and 70.3% specificity with a SCID diagnosis of major depressive disorder (Lamoureux et al., Reference Lamoureux, Linardatos, Fresco, Bartko, Logue and Milo2010) and is consistent with clinical practices for screening and triaging patients for treatment or prevention.

Analytical approach

This study aimed to elucidate treatment mechanisms, which is predicated on treatment engagement. Indeed, data from patients who never engage in treatment cannot elucidate the mechanisms of treatment. Therefore, we utilized an established intention-to-treat approach (Fergusson, Aaron, Guyatt, & Hébert, Reference Fergusson, Aaron, Guyatt and Hébert2002; Gupta, Reference Gupta2011) that included individuals who engaged in at least one session of the allocated treatment. All individuals lost to follow-up were included in the analysis, and missing data were handled with full information maximum likelihood. This ITT approach was also adapted for circumstances where the risk for type II error is high. A critical weakness of the traditional ITT approach is its high susceptibility to type II error when attrition rate is high (Fergusson et al., Reference Fergusson, Aaron, Guyatt and Hébert2002; Hollis & Campbell, Reference Hollis and Campbell1999). Because non-usage attrition is an inherent and endemic problem in Internet-based treatments (Christensen et al., Reference Christensen, Griffiths and Farrer2009; Melville, Casey, & Kavanagh, Reference Melville, Casey and Kavanagh2010; Neve, Collins, & Morgan, Reference Neve, Collins and Morgan2010) – with non-usage rates as high as 81% (Melville et al., Reference Melville, Casey and Kavanagh2010) – the traditional ITT approach would excessively increase type II error for this trial. The rationale for this adaptation is that data from those without any treatment exposure will indicate very little about the efficacy of the treatment (Gupta, Reference Gupta2011), and at worst generate erroneous parameter estimates leading to biased results.Footnote Footnote 1

The first hypothesis tested the direct effect of dCBT-I on resilience via a mixed-effects linear regression with individuals as a random intercept to account for individual differences in baseline values. The second hypothesis tested the path models using structural equation modeling. Two sets of models were conducted separately by syndrome: one with insomnia as the outcome variable of interest (see Fig. 2) and the other with depression as the outcome variable of interest (see Fig. 3). Transdiagnostic risk for insomnia and depression was treated as a latent factor in the model with sleep reactivity, stress, and rumination as indicators of latent risk. The second hypothesis protection of resilience through reducing risk factors was tested via the indirect effect of dCBT-I on outcomes at 1-year follow-up through the impact of resilience on latent risk. Full information maximum likelihood was used to produce robust parameter estimates of the indirect effects, and statistical significance was tested using a 95% confidence interval of the unstandardized coefficients (indicated as b in the results) derived from bootstrapping drawing from 1000 samples. Standardized coefficients are also reported (indicated as β in the results). Model fit for each model was determined using four indices: the χ2 fit statistic, the root mean square error of approximation (RMSEA) <0.05 (Browne & Cudeck, Reference Browne and Cudeck1992), the Tucker–Lewis index (TLI) >0.95, and the comparative fit index (CFI) >0.95 (Bentler, Reference Bentler1990; Tucker & Lewis, Reference Tucker and Lewis1973). Pre-treatment scores were included during model building and were excluded from the final model for parsimony because it did not impact results. All analyses were conducted in R (version 4.0.5) using the LAVAAN package.

Fig. 2. Resilience by group at each study timepoint. Pre-tx, pre-treatment; Post-tx, post-treatment; 1-yr fu, 1-year follow-up. Error bars indicate 95% confidence intervals. Cohen's d of improved resilience from pre-tx due to dCBT-I: at post-tx = 0.34, at 1-year follow-up = 0.36.

Fig. 3. Means of sleep reactivity, stress, and rumination as risk factors of insomnia and depression over the three assessment timepoints (pre-treatment, post-treatment, and 1-year follow-up). dCBT-I, digital cognitive behavioral therapy for insomnia; FIRST, Ford Insomnia Response to Stress Test; PTQ, Perseverative Thinking Questionnaire; Pre-tx, pre-treatment; Post-tx, post-treatment; 1-yr fu, 1-year follow-up. Error bars represent 95% confidence intervals. Cohen's d of improved risk indicators from baseline due to dCBT-I at post-treatment: d FIRST = 0.55, d stress = 0.44, d rumination = 0.32; at 1-year follow-up: d FIRST = 0.45, d stress = 0.33, d rumination = 0.21.

Sensitivity analyses were conducted in two subsamples to examine the effect of improved resilience on prevention more directly. The first sensitivity analysis examined the impact of resilience on insomnia relapse in a subsample that achieved remission at post-treatment (n = 364). Because resilience improved with insomnia remission, we opted to use change in resilience from post-treatment to 1-year follow-up to disentangle resilience from insomnia remission. The second sensitivity analysis similarly examined the impact of resilience on depression prevention in a subsample with minimal to no depression at baseline (n = 269).

Results

At baseline, individuals in both conditions exhibited clinically significant insomnia of moderate severity. The sample also showed comorbid depression symptoms as is common in insomnia (see Table 1). Individuals reported moderate levels of sleep reactivity, stress, and rumination at baseline. As was previously reported (Cheng et al., Reference Cheng, Luik, Fellman-Couture, Peterson, Joseph, Tallent and Roth2018, Reference Cheng, Kalmbach, Castelan, Murugan and Drake2020b), the dCBT-I condition resulted in acute reductions in both insomnia (10.0 ± 5.7 s.d. point decrease on the ISI) and depression symptoms (4.8 ± 5.0 s.d. point decrease on the QIDS-SR16) that were approximately twofold that in the control condition, with effects maintained at 1-year follow-up.

Table 1. Sample characteristics by treatment condition (mean ± s.d.; or %)

dCBT-I, digital cognitive behavioral therapy for insomnia; ISI, Insomnia Severity Index; QIDS-SR16, Quick Inventory of Depressive Symptomatology – 16-item self-report; BRS, Brief Resilience Scale; FIRST, Ford Insomnia Response to Stress Test.

No significant differences by condition were found.

dCBT-I improves resilience mediated through improved sleep

In support of the first hypothesis, results indicated that dCBT-I led to higher resilience at both post-treatment [t(1262.2) = 4.0, p < 0.001], and 1-year follow-up [t(1272.3) = 4.9, p < 0.001; see Fig. 2 for means and effect sizes across time points]. A post-hoc analysis confirmed that higher resilience following dCBT-I was mediated by improvements in insomnia symptom severity (indirect effect = 0.14, 95% CI 0.08–0.23; β = 0.17).

Measurement model

Before testing our second hypothesis, we examined a measurement model to ensure the variables included were related to one another before computing the full structural model. A latent factor of risk for insomnia and depression was constructed with sleep reactivity, stress, and rumination as indicators. The factor model showed strong factor loadings for each of the three indicators (sleep reactivity = 0.70; stress = 0.77; rumination = 0.77; see Fig. 3 for means and effect sizes of the indicators across all time points). Other key constructs included change in resilience from pre- to post-treatment, depression, and insomnia outcomes. We allowed latent factors and key constructs to correlate with each other in the measurement model. The measurement model fit indices represent adequate model fit for insomnia [χ2 (8, N = 658) = 13.8, CFI = 0.99; TLI = 0.99; RMSEA = 0.03] and depression [χ2 (8, N = 658) = 12.1, CFI = 0.99; TLI = 0.99; RMSEA = 0.03].

Improved resilience as a protective mechanism against insomnia

The structural model (see Fig. 4) fit indices showed good model fit to the data [χ2 (8, N = 658) = 13.8, CFI = 0.99; TLI = 0.99; RMSEA = 0.03]. Consistent with the first hypothesis, results revealed that those in the dCBT-I condition exhibited greater improvements in resilience [b = 0.23 (β = 0.34), 95% CI 0.13–0.33]. Interestingly, improvement in resilience was also a significant mediator of the post-treatment gains in insomnia severity associated with dCBT-I [b = −0.33 (β = 0.06), 95% CI −0.54 to −0.17]. dCBT-I also reduced latent transdiagnostic risk at post-treatment compared to the control condition, which was also mediated by improved resilience [b = −0.33 (β = −0.05), 95% CI −0.57 to −0.16]. Finally, the indirect effect of reduced risk through improved resilience further predicted lower insomnia severity at 1-year follow-up [b = −0.12 (β = −0.02), 95% CI −0.24 to −0.05]. An additional sensitivity analysis in a subsample that achieved post-treatment remission indicated that each point improvement in resilience following treatment reduced the odds of insomnia relapse by 76% 1 year later (OR 0.24, 95% CI 0.12–0.45).

Fig. 4. Structural equation model relating the impact of digital cognitive behavioral therapy for insomnia (dCBT-I) on insomnia severity at post-treatment (6–12 weeks after pre-treatment) and 1-year follow-up through changes in resilience and latent risk. Fit indices showed a good model fit to the data [χ2 (8, N = 658) = 13.8, CFI = 0.99; TLI = 0.99; RMSEA = 0.03].

Improved resilience as a mediator of treatment gains in depression

The structural model (see Fig. 5) fit indices for the depression model also indicated a good model fit to the data [χ2 (8, N = 658) = 13.5, CFI = 1.00; TLI = 0.99; RMSEA = 0.03]. Like the analysis with insomnia, the model with depression as the outcome variable of interest also revealed that greater improvement in resilience (i.e. from pre- to post-treatment) predicted lower post-treatment depression severity (sans sleep items) [b = −1.22 (β = −0.19), 95% CI −1.73 to −0.78]. Analyses also revealed that dCBT-I had an indirect effect on post-treatment depression severity through improved resilience [b = −0.28 (β = −0.06), 95% CI −0.47 to −0.14]. As found in the prior model, dCBT-I reduced latent transdiagnostic risk compared to the sleep education control [b = −2.19 (β = −0.32), 95% CI −3.00 to −1.39], and that reduced risk was also mediated by improved resilience [b = −0.31 (β = −0.05), 95% CI −0.55 to −0.14]. Finally, the indirect effect of reduced risk through improved resilience further predicted lower depression severity at 1-year follow-up [b = −0.15 (β = −0.03), 95% CI −0.28 to −0.07]. A sensitivity analysis in a subsample with minimal to no depression (full score) at baseline indicated that each point improvement in resilience following treatment reduced the odds of moderate to severe depression (full score) by 65% 1 year later (OR 0.35, 95% CI 0.18–0.65).

Fig. 5. Structural equation model relating the impact of digital cognitive behavioral therapy for insomnia (dCBT-I) on depression severity (sans sleep items) at post-treatment (6–12 weeks after pre-treatment) and 1-year follow-up through changes in resilience and latent risk. Fit indices showed good model fit to the data [χ2 (8, N = 658) = 12.1, CFI = 1.00; TLI = 0.99; RMSEA = 0.03].

Discussion

The overall aim of this study was to examine if insomnia treatment improved psychological resilience and test improved resilience as a candidate mechanism in the protective effects of CBT-I. Results provided support for our hypotheses that (1) psychological resilience improved with dCBT-I, and that (2) resilience protected against insomnia and depression 1-year later by reducing transdiagnostic risk (as indicated by sleep reactivity, stress, and rumination).

Importantly, this study is among the first to take a strengths-based approach as opposed to a deficits-based approach in understanding the mechanisms underlying the clinical impact of insomnia treatment. There is a plethora of research demonstrating that insomnia – especially when left untreated – is a robust risk factor for medical morbidities. However, less research has focused on how mitigating insomnia can promote health and resilience. Indeed, emergent research has suggested that insomnia treatment not only effectively removes a disease-state but may also have an ‘upward spiral’ effect in protecting against other illnesses, perhaps through improving stress reactivity and regulation. For example, treatment of insomnia not only reduces comorbid depression symptoms but also prevents future incidence of depression and relapse (Batterham et al., Reference Batterham, Christensen, Mackinnon, Gosling, Thorndike, Ritterband and Griffiths2017; Cheng et al., Reference Cheng, Kalmbach, Tallent, Joseph, Espie and Drake2019; van't Veer-Tazelaar et al., Reference van't Veer-Tazelaar, van Marwijk, van Oppen, van Hout, van der Horst, Cuijpers and Beekman2009). One mechanism may be reductions in rumination as a component that perpetuates stress (Cheng et al., Reference Cheng, Kalmbach, Castelan, Murugan and Drake2020b). Additionally, research during the COVID-19 pandemic has found that those who received dCBT-I prior to the pandemic were protected from more adverse outcomes – including lower stress – compared to those who received sleep education, despite being comparably impacted by the pandemic (Cheng et al., Reference Cheng, Casement, Kalmbach, Castelan and Drake2020a). Our results extend this research to more directly indicate that mitigating insomnia symptom severity leads to improved resilience, which protects against future adverse stressors and health outcomes.

Consistent with other findings in the literature, our results also indicate that while reducing insomnia bolsters resilience, the opposite also appears true – that reductions in insomnia associated with dCBT-I are also mediated by improved resilience (i.e. a bidirectional relationship, albeit to a smaller degree). Although dCBT-I does not explicitly target resilience (hence the smaller effect), its components may still bolster response and recovery from stress independently from improved sleep. For example, when deployed effectively, cognitive strategies taught in CBT-I (e.g. cognitive reappraisal, paradoxical intention, etc.) may help patients recover from stress more effectively. Furthermore, reductions in dysfunctional beliefs about sleep may also reduce perseverative thinking about sleepless nights, further aiding stress management and recovery. The improvement in stress management and recovery results in improved resilience, which then contributes to the acute reduction of and protection from future insomnia and depression.

Results from this study support a framework for conceptualizing how improvement in psychological resilience following dCBT-I underlies improvements in risk for insomnia and depression (related to stress reactivity) up to 1-year following treatment. Results found that those in the dCBT-I group showed a 4.9-point reduction in sleep reactivity (3.1 points more than the control group), which was partially mediated through improved resilience. Sleep reactivity is a robust predisposing risk factor for insomnia that describes a vulnerability in the sleep system to perturbations due to a range of stimuli including external/environmental (e.g. first night effect) (Chen et al., Reference Chen, Jarrin, Rochefort, Lamy, Ivers and Morin2015; Kalmbach et al., Reference Kalmbach, Anderson and Drake2018), psychological (Drake et al., Reference Drake, Friedman, Wright and Roth2011), and physiological (e.g. caffeine) (Drake et al., Reference Drake, Jefferson, Roehrs and Roth2006). Our prospective research has shown that the development of insomnia disorder involves a sensitization of sleep reactivity (an average of 3.8 points) that increases the likelihood of symptom exacerbation or recurrence from stress (Kalmbach, Pillai, Arnedt, Anderson, & Drake, Reference Kalmbach, Pillai, Arnedt, Anderson and Drake2016a). As the reduction in sleep reactivity following dCBT-I is comparable in magnitude, these findings suggest that dCBT-I may reverse this sensitization via improved resilience, thereby mitigating sleep reactivity as a contributor to recurrent insomnia.

These results also suggest that dCBT-I may reduce stress as a risk factor of insomnia and depression. Indeed, not only did we see reduced stress following dCBT-I at 1-year follow-up, a follow-up to this study revealed that the improvements in stress persisted 3–4 years following treatment (during the COVID-19) (Cheng et al., Reference Cheng, Casement, Kalmbach, Castelan and Drake2020a), suggesting that the protective effects of dCBT-I are highly durable. In conjunction with prior research, it may be that resilience buffers against a more severe stress response via appraisals of stressors as less negative or unmanageable (Morin et al., Reference Morin, Rodrigue and Ivers2003; Pillai et al., Reference Pillai, Roth, Mullins and Drake2014; Voss et al., Reference Voss, Kolling and Heidenreich2006). These results are also consistent with the literature on the importance of sleep in mood and emotion regulation, which is a critical component in stress recovery and thus resilience. Studies have shown that poor sleep is associated with heightened negative affect and blunted positive affect (Ong, Cardé, Gross, & Manber, Reference Ong, Cardé, Gross and Manber2011). Additionally, disturbed sleep also sensitizes attention to negative emotional stimuli and attenuates responsivity to positive events (Jackson et al., Reference Jackson, Gunzelmann, Whitney, Hinson, Belenky, Rabat and Van Dongen2013; Zohar, Tzischinsky, Epstein, & Lavie, Reference Zohar, Tzischinsky, Epstein and Lavie2005). Under the stress-diathesis model, this increases risk for psychopathology as it not only enhances baseline stress, but also potentiates the stress response to negative stimuli and detracts from any protective impact of positive events. Conversely, restoring sleep health appears to attenuate risk as indicated by stress, sleep reactivity, and rumination, thus promoting more effective recovery from stressors. Indeed, results indicated that improvements in resilience contributed to the prevention of depression even after accounting for improvements in rumination as another established mechanism for depression prevention (Cheng et al., Reference Cheng, Kalmbach, Castelan, Murugan and Drake2020b).

Some limitations should be considered in interpreting these results. First, psychological resilience relied on a self-report measure, albeit one that has been validated. Further research should replicate these findings with objective measures of psychological resilience. Second, individuals with severe depression with acute suicidal ideation were excluded from this study due to the aims of depression prevention; however, individuals did develop moderate to severe depression at the follow-up assessments. Finally, in alignment with the analytical approach, interpretation of parameter estimates should only be generalized to those with some exposure to dCBT-I.

Conclusion

This randomized controlled trial examined stress and resilience following dCBT-I. Results demonstrated that dCBT-I improved psychological resilience, and that this improvement was a contributing mechanism to reducing risk for insomnia and depression 1 year later. These results suggest that, in addition to mitigating the disorder, insomnia treatment may also bolster protective factors that enhance adaptation and response to stressful life events. This has implications for insomnia treatment as a candidate strategy for stress inoculation in those at risk for adverse health outcomes.

Acknowledgements

We would like to thank the Division of Sleep Medicine and the staff at the Thomas Roth Sleep Disorders and Research Center at the Henry Ford Health System for their continued support.

Financial support

This work was supported by the Robert Wood Johnson Foundation (CLD). Support for PC was provided from the National Heart Lung and Blood Institute (K23HL138166).

Conflict of interest

None.

Ethical standards

The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008.

Footnotes

Clinical Trial Registry: clinicaltrials.gov

Clinical Trial Identifier: NCT02988375

The notes appear after the main text.

1 Those with and without exposure to treatment had near zero difference in resilience (b = 0.008, s.e. = 0.124, p = 0.952).

References

American Psychiatric Association (2013). Diagnostic and statistical manual of mental disorders (DSM-5®). Arlington, VA: American Psychiatric Publishing.Google Scholar
Baglioni, C., Battagliese, G., Feige, B., Spiegelhalder, K., Nissen, C., Voderholzer, U., … Riemann, D. (2011). Insomnia as a predictor of depression: A meta-analytic evaluation of longitudinal epidemiological studies. Journal of Affective Disorders, 135(1–3), 1019. https://doi.org/10.1016/j.jad.2011.01.011.CrossRefGoogle ScholarPubMed
Ballesio, A., Ottaviani, C., & Lombardo, C. (2019). Poor cognitive inhibition predicts rumination about insomnia in a clinical sample. Behavioral Sleep Medicine, 17(5), 672681. https://doi.org/10.1080/15402002.2018.1461103.CrossRefGoogle Scholar
Barnes, C. M., Miller, J. A., & Bostock, S. (2017). Helping employees sleep well: Effects of cognitive behavioral therapy for insomnia on work outcomes. Journal of Applied Psychology, 102(1), 104113. https://doi.org/10.1037/apl0000154.CrossRefGoogle ScholarPubMed
Batterham, P. J., Christensen, H., Mackinnon, A. J., Gosling, J. A., Thorndike, F. P., Ritterband, L. M., … Griffiths, K. M. (2017). Trajectories of change and long-term outcomes in a randomised controlled trial of internet-based insomnia treatment to prevent depression. British Journal of Psychiatry Open, 3(5), 228235. https://doi.org/10.1192/bjpo.bp.117.005231.CrossRefGoogle Scholar
Bellesi, M., Bushey, D., Chini, M., Tononi, G., & Cirelli, C. (2016). Contribution of sleep to the repair of neuronal DNA double-strand breaks: Evidence from flies and mice. Scientific Reports, 6(1), 36804. https://doi.org/10.1038/srep36804.CrossRefGoogle Scholar
Bellesi, M., Haswell, J. D., de Vivo, L., Marshall, W., Roseboom, P. H., Tononi, G., … Cirelli, C. (2018). Myelin modifications after chronic sleep loss in adolescent mice. Sleep, 41(5), zsy034. https://doi.org/10.1093/sleep/zsy034.CrossRefGoogle ScholarPubMed
Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107(2), 238. https://doi.org/10.1037/0033-2909.107.2.238.CrossRefGoogle ScholarPubMed
Bernert, R. A., Turvey, C. L., Conwell, Y., & Joiner, T. E. (2014). Association of poor subjective sleep quality with risk for death by suicide during a 10-year period: A longitudinal, population-based study of late life. Journal of the American Medical Association Psychiatry, 71(10), 11291137. https://doi.org/10.1001/jamapsychiatry.2014.1126.Google ScholarPubMed
Bonanno, G. A. (2005). Resilience in the face of potential trauma. Current Directions in Psychological Science, 14(3), 135138. https://doi.org/10.1111/j.0963-7214.2005.00347.x.CrossRefGoogle Scholar
Bostock, S., Luik, A. I., & Espie, C. A. (2016). Sleep and productivity benefits of digital cognitive behavioral therapy for insomnia: A randomized controlled trial conducted in the workplace environment. Journal of Occupational and Environmental Medicine, 58(7), 683. https://doi.org/10.1097/JOM.0000000000000778.CrossRefGoogle ScholarPubMed
Brand, S., Gerber, M., Kalak, N., Kirov, R., Lemola, S., Clough, P. J., … Holsboer-Trachsler, E. (2014a). Adolescents with greater mental toughness show higher sleep efficiency, more deep sleep and fewer awakenings after sleep onset. Journal of Adolescent Health, 54(1), 109113. https://doi.org/10.1016/j.jadohealth.2013.07.017.CrossRefGoogle ScholarPubMed
Brand, S., Gerber, M., Kalak, N., Kirov, R., Lemola, S., Clough, P. J., … Holsboer-Trachsler, E. (2014b). “Sleep well, our tough heroes!” – In adolescence, greater mental toughness is related to better sleep schedules. Behavioral Sleep Medicine, 12(6), 444454. https://doi.org/10.1080/15402002.2013.825839.CrossRefGoogle ScholarPubMed
Brand, S., Kalak, N., Gerber, M., Clough, P. J., Lemola, S., Pühse, U., & Holsboer-Trachsler, E. (2016). During early and mid-adolescence, greater mental toughness is related to increased sleep quality and quality of life. Journal of Health Psychology, 21(6), 905915. https://doi.org/10.1177%2F1359105314542816.CrossRefGoogle ScholarPubMed
Breslau, N., Roth, T., Rosenthal, L., & Andreski, P. (1996). Sleep disturbance and psychiatric disorders: A longitudinal epidemiological study of young adults. Biological Psychiatry, 39(6), 411418. https://doi.org/10.1016/0006-3223(95)00188-3.CrossRefGoogle ScholarPubMed
Browne, M. W., & Cudeck, R. (1992). Alternative ways of assessing model fit. Sociological Methods & Research, 21(2), 230258. https://doi.org/10.1177%2F0049124192021002005.CrossRefGoogle Scholar
Buysse, D. J. (2014). Sleep health: Can we define it? Does it matter? Sleep, 37(1), 917. https://doi.org/10.5665/sleep.3298.CrossRefGoogle ScholarPubMed
Buysse, D. J., Reynolds, C. F. III, Kupfer, D. J., Thorpy, M. J., Bixler, E., Kales, A., … Roth, T. (1997). Effects of diagnosis on treatment recommendations in chronic insomnia – A report from the APA/NIMH DSM-IV field trial. Sleep, 20(7), 542552. https://doi.org/10.1093/sleep/20.7.542.CrossRefGoogle ScholarPubMed
Carney, C. E., Harris, A. L., Moss, T. G., & Edinger, J. D. (2010). Distinguishing rumination from worry in clinical insomnia. Behaviour Research and Therapy, 48(6), 540546. https://doi.org/10.1016/j.brat.2010.03.004.CrossRefGoogle ScholarPubMed
Carver, C. S. (1998). Resilience and thriving: Issues, models, and linkages. Journal of Social Issues, 54(2), 245266. https://psycnet.apa.org/doi/10.1111/0022-4537.641998064.CrossRefGoogle Scholar
Chen, I. Y., Jarrin, D. C., Rochefort, A., Lamy, M., Ivers, H., & Morin, C. M. (2015). Validation of the French version of the Ford insomnia response to stress test and the association between sleep reactivity and hyperarousal. Sleep Medicine, 16, S238. https://doi.org/10.1016/j.sleep.2015.02.1510.CrossRefGoogle Scholar
Cheng, P., Casement, M. D., Kalmbach, D. A., Castelan, A. C., & Drake, C. L. (2020a). Digital cognitive behavioral therapy for insomnia promotes later health resilience during the coronavirus disease 19 (COVID-19) pandemic. Sleep, 44(4), zsaa258. https://doi.org/10.1093/sleep/zsaa258.CrossRefGoogle Scholar
Cheng, P., Kalmbach, D., Tallent, G., Joseph, C., Espie, C. A., & Drake, C. (2019). Depression prevention via digital CBT for insomnia: A randomized controlled trial. Sleep, 42(10), zsz150. https://doi.org/10.1093/sleep/zsz150.CrossRefGoogle ScholarPubMed
Cheng, P., Kalmbach, D. A., Castelan, A. C., Murugan, N., & Drake, C. L. (2020b). Depression prevention in digital cognitive behavioral therapy for insomnia: Is rumination a mediator? Journal of Affective Disorders, 273, 434441. https://doi.org/10.1016/j.jad.2020.03.184.CrossRefGoogle ScholarPubMed
Cheng, P., Luik, A. I., Fellman-Couture, C., Peterson, E., Joseph, C. L., Tallent, G., … Roth, T. (2018). Efficacy of digital CBT for insomnia to reduce depression across demographic groups: A randomized trial. Psychological Medicine, 49(3), 110. https://doi.org/10.1017/S0033291718001113.Google ScholarPubMed
Christensen, H., Griffiths, K. M., & Farrer, L. (2009). Adherence in internet interventions for anxiety and depression: Systematic review. Journal of Medical Internet Research, 11(2), e13. https://doi.org/10.2196/jmir.1194.CrossRefGoogle Scholar
Dolsen, M. R., Asarnow, L. D., & Harvey, A. G. (2014). Insomnia as a transdiagnostic process in psychiatric disorders. Current Psychiatry Reports, 16(9), 471. https://doi.org/10.1007/s11920-014-0471-y.CrossRefGoogle ScholarPubMed
Drake, C. L., Cheng, P., Almeida, D. M., & Roth, T. (2017). Familial risk for insomnia is associated with abnormal cortisol response to stress. Sleep, 40(10), zsx143. https://doi.org/10.1093/sleep/zsx143.CrossRefGoogle ScholarPubMed
Drake, C. L., Friedman, N. P., Wright, K. P., & Roth, T. (2011). Sleep reactivity and insomnia: Genetic and environmental influences. Sleep, 34(9), 1179–1188. https://doi.org/10.5665/sleep.1234.CrossRefGoogle ScholarPubMed
Drake, C. L., Jefferson, C., Roehrs, T., & Roth, T. (2006). Stress-related sleep disturbance and polysomnographic response to caffeine. Sleep Medicine, 7(7), 567572. https://doi.org/10.1016/j.sleep.2006.03.019.CrossRefGoogle ScholarPubMed
Drake, C. L., Pillai, V., & Roth, T. (2014). Stress and sleep reactivity: A prospective investigation of the stress-diathesis model of insomnia. Sleep, 37(8), 12951304. https://doi.org/10.5665/sleep.3916.CrossRefGoogle ScholarPubMed
Ehring, T., Zetsche, U., Weidacker, K., Wahl, K., Schönfeld, S., & Ehlers, A. (2011). The Perseverative Thinking Questionnaire (PTQ): Validation of a content-independent measure of repetitive negative thinking. Journal of Behavior Therapy and Experimental Psychiatry, 42(2), 225232. https://doi.org/10.1016/j.jbtep.2010.12.003.CrossRefGoogle ScholarPubMed
Elo, A.-L., Leppänen, A., & Jahkola, A. (2003). Validity of a single-item measure of stress symptoms. Scandinavian Journal of Work, Environment & Health, 29(6), 444451. https://doi.org/10.5271/sjweh.752.CrossRefGoogle ScholarPubMed
Espie, C. A., Emsley, R., Kyle, S. D., Gordon, C., Drake, C. L., Siriwardena, A. N., … Luik, A. I. (2018). Effect of digital cognitive behavioral therapy for insomnia on health, psychological well-being, and sleep-related quality of life: A randomized clinical trial. Journal of the American Medical Association Psychiatry, 76(1), 21–30. https://doi.org/10.1001/jamapsychiatry.2018.2745.Google Scholar
Espie, C. A., Kyle, S. D., Williams, C., Ong, J. C., Douglas, N. J., Hames, P., & Brown, J. S. (2012). A randomized, placebo-controlled trial of online cognitive behavioral therapy for chronic insomnia disorder delivered via an automated media-rich web application. Sleep, 35(6), 769781. https://doi.org/10.5665/sleep.1872.CrossRefGoogle ScholarPubMed
Fergusson, D., Aaron, S. D., Guyatt, G., & Hébert, P. (2002). Post-randomisation exclusions: The intention to treat principle and excluding patients from analysis. British Medical Journal, 325(7365), 652654. https://doi.org/10.1136/bmj.325.7365.652.CrossRefGoogle ScholarPubMed
Ford, D., & Kamerow, D. (1989). Epidemiologic study of sleep disturbances and psychiatric disorders: An opportunity for prevention? Journal of the American Medical Association, 262(11), 14791484. https://doi.org/10.1001/jama.1989.03430110069030.CrossRefGoogle ScholarPubMed
Freeman, D., Sheaves, B., Goodwin, G. M., Yu, L.-M., Nickless, A., Harrison, P. J., … Espie, C. A. (2017). The effects of improving sleep on mental health (OASIS): A randomised controlled trial with mediation analysis. The Lancet Psychiatry, 4(10), 749758. https://doi.org/10.1016/S2215-0366(17)30328-0.CrossRefGoogle ScholarPubMed
Gupta, S. K. (2011). Intention-to-treat concept: A review. Perspectives in Clinical Research, 2(3), 109112. https://doi.org/10.4103/2229-3485.83221.CrossRefGoogle ScholarPubMed
Hall Brown, T. S., Akeeb, A., & Mellman, T. A. (2015). The role of trauma type in the risk for insomnia. Journal of Clinical Sleep Medicine, 11(7), 735739. https://doi.org/10.5664/jcsm.4846.CrossRefGoogle ScholarPubMed
Harvey, C.-J., Gehrman, P., & Espie, C. A. (2014). Who is predisposed to insomnia: A review of familial aggregation, stress-reactivity, personality and coping style. Sleep Medicine Reviews, 18(3), 237247. https://doi.org/10.1016/j.smrv.2013.11.004.CrossRefGoogle ScholarPubMed
Healey, S. E., Kales, A., Monroe, L. J., Bixler, E. O., Chamberlin, K., & Soldatos, C. R. (1981). Onset of insomnia: Role of life-stress events. Psychosomatic Medicine, 43(5), 439451. https://doi.org/10.1097/00006842-198110000-00007.CrossRefGoogle ScholarPubMed
Hollis, S., & Campbell, F. (1999). What is meant by intention to treat analysis? Survey of published randomised controlled trials. British Medical Journal, 319(7211), 670674. https://doi.org/10.1136/bmj.319.7211.670.CrossRefGoogle ScholarPubMed
Jackson, M. L., Gunzelmann, G., Whitney, P., Hinson, J. M., Belenky, G., Rabat, A., & Van Dongen, H. P. (2013). Deconstructing and reconstructing cognitive performance in sleep deprivation. Sleep Medicine Reviews, 17(3), 215225. https://doi.org/10.1016/j.smrv.2012.06.007.CrossRefGoogle ScholarPubMed
Jarrin, D. C., Chen, I. Y., Ivers, H., Drake, C. L., & Morin, C. M. (2016). Temporal stability of the Ford insomnia response to stress test (FIRST). Journal of Clinical Sleep Medicine, 12(10), 13731378. https://doi.org/10.5664/jcsm.6192.CrossRefGoogle ScholarPubMed
Kalmbach, D. A., Anderson, J. R., & Drake, C. L. (2018). The impact of stress on sleep: Pathogenic sleep reactivity as a vulnerability to insomnia and circadian disorders. Journal of Sleep Research, 27(6), e12710. https://doi.org/10.1111/jsr.12710.CrossRefGoogle ScholarPubMed
Kalmbach, D. A., Pillai, V., Arnedt, J. T., Anderson, J. R., & Drake, C. L. (2016a). Sleep system sensitization: Evidence for changing roles of etiological factors in insomnia. Sleep Medicine, 21, 6369. https://doi.org/10.1016/j.sleep.2016.02.005.CrossRefGoogle ScholarPubMed
Kalmbach, D. A., Pillai, V., Arnedt, J. T., & Drake, C. L. (2016b). DSM-5 insomnia and short sleep: Comorbidity landscape and racial disparities. Sleep, 39(12), 21012111. https://doi.org/10.5665/sleep.6306.CrossRefGoogle ScholarPubMed
Kalmbach, D. A., Pillai, V., Cheng, P., Arnedt, J. T., & Drake, C. L. (2015). Shift work disorder, depression, and anxiety in the transition to rotating shifts: The role of sleep reactivity. Sleep Medicine, 16(12), 15321538. https://doi.org/10.1016/j.sleep.2015.09.007.CrossRefGoogle ScholarPubMed
Lamoureux, B. E., Linardatos, E., Fresco, D. M., Bartko, D., Logue, E., & Milo, L. (2010). Using the QIDS-SR16 to identify major depressive disorder in primary care medical patients. Behavior Therapy, 41(3), 423431. https://doi.org/10.1016/j.beth.2009.12.002.CrossRefGoogle ScholarPubMed
Li, L., Wu, C., Gan, Y., Qu, X., & Lu, Z. (2016). Insomnia and the risk of depression: A meta-analysis of prospective cohort studies. BioMed Central Psychiatry, 16(1), 375. https://doi.org/10.1186/s12888-016-1075-3.CrossRefGoogle ScholarPubMed
Luo, C., Zhang, J., & Pan, J. (2013). One-year course and effects of insomnia in rural Chinese adolescents. Sleep, 36(3), 377384. https://doi.org/10.5665/sleep.2454.CrossRefGoogle ScholarPubMed
Malmberg, M., & Larsen, J. K. (2015). Depressive symptoms: The interaction between rumination and self-reported insomnia. Depression Research and Treatment, 2015, 150828. https://doi.org/10.1155/2015/150828.CrossRefGoogle ScholarPubMed
McGrath, E. R., Espie, C. A., Power, A., Murphy, A. W., Newell, J., Kelly, C., … O'Donnell, M. J. (2017). Sleep to lower elevated blood pressure: A randomized controlled trial (SLEPT). American Journal of Hypertension, 30(3), 319327. https://doi.org/10.1093/ajh/hpw132.CrossRefGoogle ScholarPubMed
Melville, K. M., Casey, L. M., & Kavanagh, D. J. (2010). Dropout from Internet-based treatment for psychological disorders. British Journal of Clinical Psychology, 49(4), 455471. https://doi.org/10.1348/014466509X472138.CrossRefGoogle ScholarPubMed
Morgenthaler, T., Kramer, M., Alessi, C., Friedman, L., Boehlecke, B., & Brown, T., … Swick, T. (2006). Practice parameters for the psychological and behavioral treatment of insomnia: An update. An American Academy of Sleep Medicine report. Sleep, 29(11), 1415. https://doi.org/10.1093/sleep/29.11.1415.Google ScholarPubMed
Morin, C. M., Belleville, G., Bélanger, L., & Ivers, H. (2011). The Insomnia Severity Index: Psychometric indicators to detect insomnia cases and evaluate treatment response. Sleep, 34(5), 601608. https://doi.org/10.1093/sleep/34.5.601.CrossRefGoogle ScholarPubMed
Morin, C. M., Rodrigue, S., & Ivers, H. (2003). Role of stress, arousal, and coping skills in primary insomnia. Psychosomatic Medicine, 65(2), 259267. https://psycnet.apa.org/doi/10.1097/01.PSY.0000030391.09558.A3.CrossRefGoogle ScholarPubMed
National Heart Lung and Blood Institute. (2011). Your guide to healthy sleep. National Institutes of Health. Retrieved from www.nhlbi.nih.gov/health-topics/all-publications-and-resources/your-guide-healthy-sleep.Google Scholar
Neve, M. J., Collins, C. E., & Morgan, P. J. (2010). Dropout, nonusage attrition, and pretreatment predictors of nonusage attrition in a commercial web-based weight loss program. Journal of Medical Internet Research, 12(4), e1640. https://doi.org/10.2196/jmir.1640.CrossRefGoogle Scholar
Nolen-Hoeksema, S. (1991). Responses to depression and their effects on the duration of depressive episodes. Journal of Abnormal Psychology, 100(4), 569. https://doi.org/10.1037//0021-843x.100.4.569.CrossRefGoogle ScholarPubMed
Ohayon, M. M. (2002). Epidemiology of insomnia: What we know and what we still need to learn. Sleep Medicine Reviews, 6(2), 97111. https://doi.org/10.1053/smrv.2002.0186.CrossRefGoogle ScholarPubMed
Ong, J. C., Cardé, N. B., Gross, J. J., & Manber, R. (2011). A two-dimensional approach to assessing affective states in good and poor sleepers. Journal of Sleep Research, 20(4), 606610. https://doi.org/10.1111/j.1365-2869.2011.00907.x.CrossRefGoogle ScholarPubMed
Palagini, L., Cipollone, G., Masci, I., Novi, M., Caruso, D., Kalmbach, D. A., & Drake, C. L. (2019). Stress-related sleep reactivity is associated with insomnia, psychopathology and suicidality in pregnant women: Preliminary results. Sleep Medicine, 56, 145150. https://doi.org/10.1016/j.sleep.2019.01.009.CrossRefGoogle ScholarPubMed
Palagini, L., Moretto, U., Novi, M., Masci, I., Caruso, D., Drake, C. L., & Riemann, D. (2018). Lack of resilience is related to stress-related sleep reactivity, hyperarousal, and emotion dysregulation in insomnia disorder. Journal of Clinical Sleep Medicine, 14(5), 759766. https://doi.org/10.5664/jcsm.7100.CrossRefGoogle ScholarPubMed
Pigeon, W. R., Bishop, T. M., & Krueger, K. M. (2017). Insomnia as a precipitating factor in new onset mental illness: A systematic review of recent findings. Current Psychiatry Reports, 19(8), 44. https://doi.org/10.1007/s11920-017-0802-x.CrossRefGoogle ScholarPubMed
Pillai, V., Anderson, J. R., Cheng, P., Bazan, L., Bostock, S., Espie, C. A., … Drake, L. (2015). The anxiolytic effects of cognitive behavior therapy for insomnia: Preliminary results from a web-delivered protocol. Journal of Sleep Medicine and Disorders, 2(2), 7. Retrieved from https://pubmed.ncbi.nlm.nih.gov/32195356/.Google ScholarPubMed
Pillai, V., Roth, T., Mullins, H. M., & Drake, C. L. (2014). Moderators and mediators of the relationship between stress and insomnia: Stressor chronicity, cognitive intrusion, and coping. Sleep, 37(7), 11991208. https://doi.org/10.5665/sleep.3838.CrossRefGoogle ScholarPubMed
Qaseem, A., Kansagara, D., Forciea, M. A., Cooke, M., & Denberg, T. D. (2016). Management of chronic insomnia disorder in adults: A clinical practice guideline from the American College of Physicians Management of Chronic Insomnia Disorder in Adults. Annals of Internal Medicine, 165(2), 125133. https://doi.org/10.7326/M15-2175.CrossRefGoogle ScholarPubMed
Riemann, D., Baglioni, C., Bassetti, C., Bjorvatn, B., Dolenc Groselj, L., Ellis, J. G., … Spiegelhalder, K. (2017). European guideline for the diagnosis and treatment of insomnia. Journal of Sleep Research, 26(6), 675700. https://doi.org/10.1111/jsr.12594.CrossRefGoogle ScholarPubMed
Roth, T. (2007). Insomnia: Definition, prevalence, etiology, and consequences. Journal of Clinical Sleep Medicine, 3(Suppl. 5), S7. https://doi.org/10.5664/jcsm.26929.CrossRefGoogle ScholarPubMed
Roth, T., & Roehrs, T. (2003). Insomnia: Epidemiology, characteristics, and consequences. Clinical Cornerstone, 5(3), 515. https://doi.org/10.1016/S1098-3597(03)90031-7.CrossRefGoogle ScholarPubMed
Salminen, S., Kouvonen, A., Koskinen, A., Joensuu, M., & Väänänen, A. (2014). Is a single item stress measure independently associated with subsequent severe injury: A prospective cohort study of 16,385 forest industry employees. BioMed Central Public Health, 14(1), 543. http://dx.doi.org/10.1186/1471-2458-14-543.CrossRefGoogle Scholar
Smith, B. W., Dalen, J., Wiggins, K., Tooley, E., Christopher, P., & Bernard, J. (2008). The brief resilience scale: Assessing the ability to bounce back. International Journal of Behavioral Medicine, 15(3), 194200. https://psycnet.apa.org/doi/10.1080/10705500802222972.CrossRefGoogle ScholarPubMed
Spielman, A. J., Caruso, L. S., & Glovinsky, P. B. (1987). A behavioral perspective on insomnia treatment. The Psychiatric Clinics of North America, 10(4), 541553. https://doi.org/10.1016/S0193-953X(18)30532-X.CrossRefGoogle ScholarPubMed
Surís, A., Holder, N., Holliday, R., & Clem, M. (2016). Psychometric validation of the 16 item Quick Inventory of Depressive Symptomatology Self-Report Version (QIDS-SR16) in military veterans with PTSD. Journal of Affective Disorders, 202, 1622. https://doi.org/10.1016/j.jad.2016.05.029.CrossRefGoogle ScholarPubMed
Taylor, D. J., Lichstein, K. L., & Durrence, H. H. (2003). Insomnia as a health risk factor. Behavioral Sleep Medicine, 1(4), 227247. https://doi.org/10.1207/S15402010BSM0104_5.CrossRefGoogle ScholarPubMed
Taylor, D. J., Mallory, L. J., Lichstein, K. L., Durrence, H., Riedel, B. W., & Bush, A. J. (2007). Comorbidity of chronic insomnia with medical problems. Sleep, 30(2), 213. https://doi.org/10.1093/sleep/30.2.213.CrossRefGoogle ScholarPubMed
Thorndike, F. P., Ritterband, L. M., Saylor, D. K., Magee, J. C., Gonder-Frederick, L. A., & Morin, C. M. (2011). Validation of the insomnia severity index as a web-based measure. Behavioral Sleep Medicine, 9(4), 216223. https://doi.org/10.1080/15402002.2011.606766.CrossRefGoogle ScholarPubMed
Tucker, L. R., & Lewis, C. (1973). A reliability coefficient for maximum likelihood factor analysis. Psychometrika, 38(1), 110. https://psycnet.apa.org/doi/10.1007/BF02291170.CrossRefGoogle Scholar
Tusaie, K., & Dyer, J. (2004). Resilience: A historical review of the construct. Holistic Nursing Practice, 18(1), 310. https://doi.org/10.1097/00004650-200401000-00002.CrossRefGoogle ScholarPubMed
van't Veer-Tazelaar, P. J., van Marwijk, H. W., van Oppen, P., van Hout, H. P., van der Horst, H. E., Cuijpers, P., … Beekman, A. T. (2009). Stepped-care prevention of anxiety and depression in late life: A randomized controlled trial. Archives of General Psychiatry, 66(3), 297304. https://doi.org/10.1001/archgenpsychiatry.2008.555.CrossRefGoogle ScholarPubMed
Voss, U., Kolling, T., & Heidenreich, T. (2006). Role of monitoring and blunting coping styles in primary insomnia. Psychosomatic Medicine, 68(1), 110115. https://doi.org/10.1097/01.psy.0000195881.01317.d9.CrossRefGoogle ScholarPubMed
Winer, J. R., Mander, B. A., Kumar, S., Reed, M., Baker, S. L., Jagust, W. J., & Walker, M. P. (2020). Sleep disturbance forecasts β-amyloid accumulation across subsequent years. Current Biology, 30(21), 42914298.e3. https://doi.org/10.1016/j.cub.2020.08.017.CrossRefGoogle ScholarPubMed
Yeung, A., Feldman, G., Pedrelli, P., Hails, K., Fava, M., Reyes, T., & Mundt, J. C. (2012). The Quick Inventory of Depressive Symptomatology, clinician rated and self-report: A psychometric assessment in Chinese Americans with major depressive disorder. The Journal of Nervous and Mental Disease, 200(8), 712. https://dx.doi.org/10.1097%2FNMD.0b013e318261413d.CrossRefGoogle ScholarPubMed
Zohar, D., Tzischinsky, O., Epstein, R., & Lavie, P. (2005). The effects of sleep loss on medical residents’ emotional reactions to work events: A cognitive-energy model. Sleep, 28(1), 4754. https://doi.org/10.1093/sleep/28.1.47.CrossRefGoogle ScholarPubMed
Figure 0

Fig. 1. Enrollment flow chart.

Figure 1

Fig. 2. Resilience by group at each study timepoint. Pre-tx, pre-treatment; Post-tx, post-treatment; 1-yr fu, 1-year follow-up. Error bars indicate 95% confidence intervals. Cohen's d of improved resilience from pre-tx due to dCBT-I: at post-tx = 0.34, at 1-year follow-up = 0.36.

Figure 2

Fig. 3. Means of sleep reactivity, stress, and rumination as risk factors of insomnia and depression over the three assessment timepoints (pre-treatment, post-treatment, and 1-year follow-up). dCBT-I, digital cognitive behavioral therapy for insomnia; FIRST, Ford Insomnia Response to Stress Test; PTQ, Perseverative Thinking Questionnaire; Pre-tx, pre-treatment; Post-tx, post-treatment; 1-yr fu, 1-year follow-up. Error bars represent 95% confidence intervals. Cohen's d of improved risk indicators from baseline due to dCBT-I at post-treatment: dFIRST = 0.55, dstress = 0.44, drumination = 0.32; at 1-year follow-up: dFIRST = 0.45, dstress = 0.33, drumination = 0.21.

Figure 3

Table 1. Sample characteristics by treatment condition (mean ± s.d.; or %)

Figure 4

Fig. 4. Structural equation model relating the impact of digital cognitive behavioral therapy for insomnia (dCBT-I) on insomnia severity at post-treatment (6–12 weeks after pre-treatment) and 1-year follow-up through changes in resilience and latent risk. Fit indices showed a good model fit to the data [χ2 (8, N = 658) = 13.8, CFI = 0.99; TLI = 0.99; RMSEA = 0.03].

Figure 5

Fig. 5. Structural equation model relating the impact of digital cognitive behavioral therapy for insomnia (dCBT-I) on depression severity (sans sleep items) at post-treatment (6–12 weeks after pre-treatment) and 1-year follow-up through changes in resilience and latent risk. Fit indices showed good model fit to the data [χ2 (8, N = 658) = 12.1, CFI = 1.00; TLI = 0.99; RMSEA = 0.03].