Internet-Based Cognitive Behavioral Therapy for Chronic Stress

Author(s): Elin Lindsäter, Erland Axelsson, Sigrid Salomonsson, Fredrik Santoft, Kersti
Ejeby, Brjánn Ljótsson, Torbjörn Åkerstedt, Mats Lekander and Erik Hedman-Lagerlöf
Source: Psychotherapy and Psychosomatics , 2018, Vol. 87, No. 5 (2018), pp. 296-305
Published by: S. Karger AG

Stable URL: https://www.jstor.org/stable/10.2307/48516566

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/10.2307/48516566?seq=1&cid=pdf-
reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

S. Karger AG is collaborating with JSTOR to digitize, preserve and extend access to Psychotherapy
and Psychosomatics

This content downloaded from

13.234.121.200 on Fri, 15 Sep 2023 13:48:34 +00:00
All use subject to https://about.jstor.org/terms
 Regular Article

Psychother Psychosom 2018;87:296–305 Received: December 2, 2017

Accepted after revision: June 5, 2018
DOI: 10.1159/000490742 Published online: July 24, 2018

Internet-Based Cognitive Behavioral

Therapy for Chronic Stress:
A Randomized Controlled Trial
Elin Lindsäter a Erland Axelsson a Sigrid Salomonsson c Fredrik Santoft a
Kersti Ejeby b Brjánn Ljótsson a, c Torbjörn Åkerstedt a, d Mats Lekander a, d, e
Erik Hedman-Lagerlöf a, e
a Division of Psychology, Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden; b Division

of Family Medicine, Department of Neurobiology, Care Sciences and Society (NVS), H1, Karolinska Institutet,
Stockholm, Sweden; c Centre for Psychiatry Research, Department of Clinical Neuroscience, Karolinska Institutet,
Stockholm, Sweden; d Stress Research Institute, Stockholm University, Stockholm, Sweden; e Osher Center for
Integrative Medicine, Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden

Keywords low-up. The study was preregistered at Clinicaltrials.gov:

Chronic stress · Adjustment disorder · Exhaustion disorder ·
Cognitive behavioral therapy · Internet · Randomized
controlled trial
Abstract
Background: Prolonged exposure to stress can lead to sub-
stantial suffering, impairment and societal costs. However,
access to psychological treatment is limited. Internet-based
cognitive behavioral therapy (ICBT) can be effective in re-
ducing symptoms of stress, but little is known of its effects in
clinical samples. The aim of this study was to investigate the
efficacy of ICBT for patients suffering from chronic stress, op-
erationalized as adjustment disorder (AD) and exhaustion
disorder (ED). Methods: A total of 100 adults diagnosed with
AD or ED were randomly assigned to a 12-week ICBT (n = 50)
or waitlist control condition (n = 50). Primary outcome was
the level of perceived stress (PSS). Secondary outcomes in-
cluded several mental health symptom domains as well as
functional impairment and work ability. All outcomes were
assessed at baseline, after treatment and at the 6-month fol-
NCT02540317. Results: Compared to the control condition,
patients in the ICBT group made large and significant im-
provements on the PSS (d = 1.09) and moderate to large im-
provements in secondary symptom domains. Effects were
maintained at the 6-month follow-up. There was no signifi-
cant between-group effect on functional impairment or
work ability. Conclusions: A relatively short ICBT is indicated
to be effective in reducing stress-related symptoms in a clin-
ical sample of patients with AD and ED, and has the potential
to substantially increase treatment accessibility. Results
must be replicated, and further research is needed to under-
stand the relationship between symptom reduction, func-
tional impairment and work ability. © 2018 S. Karger AG, Basel
Introduction
Prolonged or repeated exposure to stress without suf-
ficient recovery can contribute to physical and psychiatric
disease [1], and is associated with a wide range of symp-

© 2018 S. Karger AG, Basel Elin Lindsäter

Gustavsberg Primary Care Clinic
Odelbergs väg 19
E-Mail karger@karger.com
SE–134 40 Gustavsberg (Sweden)
www.karger.com/pps E-Mail elin.lindsater @ ki.se

This content downloaded from

13.234.121.200 on Fri, 15 Sep 2023 13:48:34 +00:00
All use subject to https://about.jstor.org/terms
toms including disturbed sleep, depressed mood, cogni-
tive dysfunction and somatoform complaints [1, 2]. Fur-
ther, stress leads to increased health care utilization and
high societal costs due to staff turnover, loss of productiv-
ity and sick leave [3, 4]. Cognitive behavioral therapy
(CBT) can be effective in reducing stress [5, 6], but the
availability of psychological treatment is limited [7]. One
way of addressing this problem is to deliver treatment
via the Internet. Therapist-guided Internet-based CBT
(ICBT) has been shown to be efficacious for a wide range
of clinical conditions in over 100 randomized controlled
trials [8], with important advantages including indepen-
dence of geographical distance and high scalability. A re-
cent meta-analysis indicated that ICBT can be efficacious
in reducing stress with small to moderate effect sizes [9],
and some individual studies have yielded large effects on
symptoms of stress and emotional exhaustion [10–12].
To date, however, trials have almost exclusively used sub-
clinical or undiagnosed samples, recruited from occupa-
tional settings. This limits generalizability of results to
clinical populations for whom the need for efficient and
highly available treatments is urgent. We know of only 1
study that has investigated the efficacy of ICBT in a sam-
ple where all participants were diagnosed with a stress-
related disorder (in this case according to the ICD-10
F43). That study showed that ICBT had moderate to large
effects on symptoms of stress and secondary mental
health outcomes compared to an attention control group
[12]. Although promising, that study used a sample of
distressed managers rather than investigating the efficacy
of ICBT for a broader clinical population. Additional
studies are needed to determine the efficacy of ICBT for
clinical samples suffering from chronic stress.
The aim of this study was to investigate the efficacy of
a 12-week therapist-guided ICBT protocol, compared to
a waitlist control group (WLC) for patients suffering
from chronic stress. There is a lack of international con-
sistency regarding how to define and classify clinically
significant reactions to chronic stress [2, 13]. We opera-
tionalized symptoms using diagnostic criteria for adjust-
ment disorder (AD) and exhaustion disorder (ED). AD is
a condition strongly tied to acute and chronic stress [14].
It entails clinically significant emotional or behavioral
changes in response to stressful life events that occur
within 3 months of the onset of symptoms [15]. AD is
characterized by marked distress and significant impair-
ment in social or occupational functioning. Although AD
has been criticized due to vague diagnostic criteria and
large symptom overlap both with normal reactions to
stress and subthreshold manifestations of mood and anx-
Internet-Based Cognitive Behavioral
Therapy for Chronic Stress
This content downloaded from
13.234.121.200 on Fri, 15 Sep 2023 13:48:34 +00:00
All use subject to https://about.jstor.org/terms
iety disorders [e.g., 14, 16], it is a common diagnosis in
clinical practice and merits clinical attention [14]. ED is a
more chronic and debilitating condition, characterized
by prolonged fatigue, disturbed sleep, cognitive dysfunc-
tion and somatoform complaints that are assumed to re-
sult from exposure to stressful life events in the past 6
months [2]. ED is recognized as a medical diagnosis by
the Swedish National Board of Health and Welfare and is
listed as a specification of F42.8 (other reactions to severe
stress) [17] in the Swedish version of the ICD-10. Diag-
nostic criteria for ED overlap with criteria for allostatic
overload as defined in Diagnostic Criteria for Psychoso-
matic Research [18] as well as with clinical burnout [19].
Although AD and ED differ in symptom severity and
chronicity of stress exposure, both are based on the lon-
gitudinal course of symptoms and behavioral changes in
the context of stressful life events. We therefore predicted
that these stress-related conditions could be treated with
the same protocol. Our primary hypothesis was that ICBT
would be superior to the WLC in reducing symptoms of
perceived stress. We also hypothesized that ICBT would
lead to improvement in symptoms of exhaustion, de-
pressed mood, disturbed sleep and anxiety, as well as in
sickness behavior, somatoform complaints, self-rated
health and quality of life compared to WLC. Finally, we
investigated whether ICBT would have an effect on func-
tional impairment and self-rated work ability.
Methods
Design
This was a randomized controlled efficacy trial in which 100
patients diagnosed with AD (n = 53) or ED (n = 47) were random-
ized to ICBT (n = 50) or a WLC group (n = 50). Sample size was
determined through a priori power analysis for conventional para-
metric tests to achieve 80% power to detect a between-group dif-
ference of d = 0.6, given α = 0.05 (2-tailed). The trial was conduct-
ed between September 2015 and August 2016 at Karolinska Insti-
tutet, Stockholm, Sweden, and was preregistered at Clinicaltrials.
gov (No. NCT02540317). All participants completed assessments
at baseline prior to randomization. After inclusion, a person not
related to the study randomized participants using a true random
number generator (www.random.org). Randomization was strati-
fied based on diagnosis (AD or ED) to make it possible to explore
effects for patients with different symptom severity. To reduce
waiting time for patients after inclusion, treatment start occurred
in 3 cohorts from September to November 2015. Assessments took
place at baseline, after treatment and at the 6-month follow-up
(6MFU). For ethical reasons, patients in the WLC group were
crossed over to treatment after post-treatment assessments. Hence,
6MFU data are provided for the ICBT condition only. All patients
provided written informed consent, and the regional ethics review
board in Stockholm approved the study.
Psychother Psychosom 2018;87:296–305 297
DOI: 10.1159/000490742
 Table 1. Baseline characteristics of included patients

Total sample ICBT WLC

(n = 100) (n = 50) (n = 50)

Gender female, n (%) 85 (85.0) 41 (82.0) 44 (88.0)

Age, years
Mean ± SD 46.2±8.8 45.1±8.8 47.2±8.7
Minimum–maximum 26–65 29–64 26–65
Highest education, n (%)
College/university >3 years 73 (73.0) 36 (72.0) 37 (74.0)
College/university ≤3 years 18 (18.0) 8 (16.0) 10 (20.0)
Secondary school 2–3 years 9 (9.0) 6 (12.0) 3 (6.0)
Employment, n (%)
Full time 71 (71.0) 36 (72.0) 35 (70.0)
Part time (25–75%) 10 (10.0) 5 (10.0) 5 (10.0)
Other1 5 (5.0) 3 (6.0) 2 (4.0)
Sick leave/disability pension 14 (14.0) 6 (12.0) 8 (16.0)
Medication, n (%)
Antidepressants 11 (11.0) 5 (10.0) 6 (12.0)
Anxiolytics 2 (2.0) 1 (2.0) 1 (2.0)
Hypnotics 17 (17.0) 9 (18.0) 8 (16.0)
Mean symptom duration ± SD, years 1.6±1.3 1.5±1.2 1.6±1.4
Psychiatric comorbidity, n (%) 28 (28.0) 15 (30.0) 13 (26.0)
Depression 13 (13.0) 8 (16.0) 5 (10.0)
Panic disorder 4 (4.0) 2 (4.0) 2 (4.0)
Generalized anxiety disorder 9 (9.0) 3 (6.0) 6 (12.0)
Social anxiety disorder 6 (6.0) 3 (6.0) 3 (6.0)

ICBT, Internet-based cognitive behavioral therapy; WLC, waitlist control. 1 Other employment refers to, e.g.,
parental leave or unemployment.

Recruitment, Inclusion Criteria and Participants
The study was advertised in a national newspaper and via so-
cial media. Applicants were self-referred via a study web portal,
and eligibility criteria were (a) age 18–65 years, (b) a primary di-
agnosis of AD or ED, (c) regular access to a computer and to the
Internet, (d) being able to read and write in Swedish, (e) no sub-
stance abuse or dependence in the past 6 months, (f) no current
or past psychosis or bipolar disorder, (g) no suicidal ideation, (h)
if on medication with a monoamine agonist, this had been stable
in the past month, (i) no ongoing psychological treatment, (j) no
CBT for stress-related difficulties in the past year and (k) not be-
ing abroad for more than 2 weeks during the planned treatment
period.
All applicants underwent a 60-min assessment interview with
a licensed psychologist. DSM-5 AD and ICD-10 ED were assessed
with a clinical interview developed specifically for the trial that
closely followed the diagnostic criteria for the diagnoses (online
suppl. Table 1 specifies criteria for ED; for all online suppl. mate-
rial, see www.karger.com/doi/10.1159/000490742). Comorbid
psychiatric disorders were surveyed using the Mini-International
Neuropsychiatric Interview (MINI), which is a reliable and valid
instrument for assessing psychiatric disorders [20]. Mild to mod-
erate forms of psychiatric diagnoses according to the DSM-5 were
accepted as comorbid conditions, as long as these were assessed to
be secondary to AD or ED. Because of the national recruitment,
interviews were conducted via telephone, which has shown good
agreement with face-to-face assessments of depression and anxiety
disorders [21]. Patients were given information about the study,
and they were instructed to keep their medication stable as well as
not to engage in other psychological treatments during the course
of the study.
Table 1 presents characteristics of included patients. Patient
characteristics for AD and ED, respectively, can be found in the
online supplementary Table 2. Patients with ED consistently re-
ported more severe psychiatric symptoms and functional disabil-
ity, lower quality of life, higher psychiatric comorbidity and high-
er frequency of sick leave than did patients with AD.
Interventions
The ICBT protocol was a 12-week treatment based on a CBT
protocol for AD and ED that was developed in clinical practice
and evaluated in 2 randomized clinical trials conducted by our
research group [22, 23]. A key assumption of the treatment was
that stress is a natural and necessary part of life, but that negative
health consequences result, in part, from maladaptive behaviors
adopted in response to stressors. For example, many patients
withdraw from social activities and resort to passivity in an at-
tempt to avoid stress-related symptoms. Others are in a constant
state of “hurry” and down-prioritize physical activity, leisure
time and recovery between periods of high stress. Through con-

298 Psychother Psychosom 2018;87:296–305 Lindsäter et al.

DOI: 10.1159/000490742

This content downloaded from

13.234.121.200 on Fri, 15 Sep 2023 13:48:34 +00:00
All use subject to https://about.jstor.org/terms
 Module Theme Main exercises

1 Introduction to CBT and stress-

psychoeducation Daily
activities
2 Learning recovery techniques and
diary
practicing relaxation
3 Dealing with disturbed sleep Functional
analysis
4 Identifying values and breaking them
down into concrete behaviors/steps
5 Behavioral activation, taking steps
Scheduled
6 Exposure, dealing with difficult recovery
emotions, breaking fear-avoidance
patterns Practicing
7 Continued behavior change toward life relaxation
Behavioral
balance and health
activation
8 Continued behavior change toward life in line with
balance and health values
9 Identifying core assumptions and
challenging thoughts Exposure

10 Communication skills, assertiveness

training
11 A summary of the treatment

12 Maintaining gains and relapse

prevention
CBT, Cognitive behavioral therapy

Fig. 1. Description of treatment content.

tinuous self-monitoring and functional analyses, patients identi-
fied their own cognitive and behavioral patterns believed to
maintain and exacerbate their symptoms. Contemporary CBT
techniques such as sleep management [24], behavioral activation
[25] and exposure [26] (e.g., breaking fear-avoidance patterns
related to assertiveness, perfectionism and excessive worry) were
used to increase recovery and establish health-promoting activi-
ties in line with patients’ values. Figure 1 provides an overview of
treatment content.
The treatment content was conveyed through an online self-
help text, divided into 12 modules similar to chapters of a book.
All modules included exercises and homework assignments. Pa-
tients were encouraged to complete approximately 1 module per
week, and the duration of the treatment was 12 weeks. Modules
were made available in a consecutive order, and patients had to
complete 1 module before given access to the next. Each patient
was guided through the treatment by a therapist via a secure asyn-
chronous online contact system. Because the treatment content
was identically presented to each patient and included all psycho-
education and instructions for behavioral change, the role of the
therapist was limited to providing feedback on homework assign-
ments and giving emotional and technical support. At any time,
patients could send messages with questions or concerns to the
therapist, and expect an answer within 48 h. Patients additionally
received weekly, automated SMS reminders to log on to the treat-
ment platform and work with the treatment. The therapists were
2 licensed clinical psychologists with experience of working with
CBT and stress in primary care settings.
Patients in the WLC condition were given no treatment during
the 12-week main phase of the trial. They were, however, instruct-
ed to fill in the same weekly questionnaires (see “Outcome Mea-
sures”) as the ICBT group and received automated SMS reminders
to do so.
Outcome Measures
The primary outcome measure was the 14-item Perceived
Stress Scale (PSS [27]), measuring how often one has perceived life
as unpredictable, uncontrollable and overloading in the past
month. The total scale range is 0–56 where a higher score indicates
more stress. The Swedish version of the PSS-14, used in the present
study, has shown good internal consistency, construct validity and
sensitivity to change [28].
Secondary outcome measures included the 22-item Shirom-
Melamed Burnout Questionnaire (SMBQ [29]), which was used
to assess symptoms of exhaustion. We used the Montgomery-
Åsberg Depression Rating Scale Self-Report (MADRS-S [30]), the

Internet-Based Cognitive Behavioral Psychother Psychosom 2018;87:296–305 299

Therapy for Chronic Stress DOI: 10.1159/000490742

This content downloaded from

13.234.121.200 on Fri, 15 Sep 2023 13:48:34 +00:00
All use subject to https://about.jstor.org/terms
 173 individuals applied to participate and were assessed for eligibility

Excluded, n = 73
Not primary AD or ED, n = 28
Ongoing psychological treatment for stress, n = 7
CBT for stress in the past year, n = 16
Not stable medication, n = 2
>2 weeks abroad, n = 3
Bipolar disorder, n = 2
Substance abuse, n = 1
No longer interested, n = 14

100 individuals were included in the study, completed pretreatment assessment and
underwent randomization

ICBT, n = 50 WLC, n = 50
Received intended treatment,
n = 41 (82 %)1

Fig. 2. Participant flow and reasons for Completed post-treatment Completed post-treatment
dropping out throughout the trial. AD, ad- assessment, n = 49 (98%) assessment,
justment disorder; ED, exhaustion disor- Completed 6MFU, n = 48 (96%)
der; ICBT, Internet-based cognitive behav- n = 47 (94%)
ioral therapy; WLC, waitlist control. 1 Not
receiving intended treatment was defined
as completing < 50% of the modules. The Included in intention-to-treat Included in intention-to-treat
main reason for non-completion was that analysis, n =50 analysis, n = 50
the treatment was found to be too compre-
hensive.

Insomnia Severity Index (ISI [31]) and the 7-item Generalized
Anxiety Disorder Scale (GAD-7 [32]) to assess depression, insom-
nia and anxiety, respectively. We further assessed sickness behav-
ior (SicknessQ [33]), self-rated health (SRH-5 [34]), somatoform
complaints (PHQ-15 [35]) and quality of life (BBQ [36]). Lastly,
functional impairment (WHODAS 2.0 [37]) and work ability
(WAI [38]) were measured.
All primary and secondary outcome measures were self-report
questionnaires filled in via the online study platform and admin-
istered before and after treatment (the 12-week trial phase), as well
as (for the ICBT group) 6 months after treatment. PSS and SMBQ
were also administered on a weekly basis. Some of the data that
were collected in the trial (e.g., mediators, health economic out-
comes; see Clinicaltrials.gov) will be reported elsewhere.
To assess treatment credibility, the C-Scale [39] was adminis-
tered to patients in the ICBT group in weeks 2 and 8. The Client
Satisfaction Questionnaire-8 [40] was used to assess satisfaction
with the treatment and was administered to the ICBT group after
treatment, together with questions about to which extent they
worked with the treatment and took part of the different modules.
Adverse events were assessed (in the ICBT group) after treatment
and at 6MFU through an item where patients were asked to state
whether they experienced any negative or unwanted effects of the
treatment. It they answered “Yes”, they were asked to specify the
adverse event in free text.
Statistical Analyses
Data analysis was conducted using SPSS 20.0 (IBM, Chicago).
An intention-to-treat approach was used, meaning that all patients
who were randomized were included in the analysis. Continuous
outcomes were analyzed by means of mixed-effect modeling, tak-
ing individual heterogeneity into account [41]. A random inter-
cept was used for all analyses and, whenever it improved the mod-
el, a random slope. Model fit was determined using –2 restricted
log likelihoods. The main statistical parameter of interest was the
interaction effect of time (pre- to post-treatment) and study group
(ICBT vs. WLC). Subgroup analyses were conducted for AD and
ED, respectively, and the effect of diagnostic group on treatment
outcome was tested by adding a 3-way interaction effect between
time, study group and diagnostic group to the mixed model. Treat-
ment credibility, treatment satisfaction and adverse events were
analyzed with t tests, and nominal data were analyzed with χ2 tests
or Fisher’s exact test. Within- and between-group effect sizes were
calculated using Cohen’s d, and the proportion of patients with
clinically significant improvement on the outcomes PSS and
SMBQ were calculated using the criteria proposed by Jacobson and

300 Psychother Psychosom 2018;87:296–305 Lindsäter et al.

DOI: 10.1159/000490742

This content downloaded from

13.234.121.200 on Fri, 15 Sep 2023 13:48:34 +00:00
All use subject to https://about.jstor.org/terms
 Primary Outcome Measures

Color version available online

45 Figure 3 shows the improvement on the primary out-
ICBT mean
40 come measure PSS. Table 2 presents means, standard de-
WLC mean viations, interaction effects and effect sizes on PSS.
35
30 There was a superior reduction in perceived stress in
25 the ICBT group compared to WLC. The between-group
PSS

20 effect size at the post-treatment assessment was large in

15
favor of ICBT. There was no significant difference be-
10
tween post-treatment and 6MFU findings (F = 2.4; df =
5
1, 46; p = 0.128), indicating stability of treatment gains.
0
Pre Post 6MFU Secondary Outcome Measures
Assessment points Table 2 shows means, standard deviations, interaction
effects and effect sizes for continuous secondary out-
comes. Due to space limitations, outcomes regarding
Fig. 3. Course of improvement on the primary outcome measure sickness behavior, somatoform complaints, self-rated
Perceived Stress Scale (PSS). Error bars represent 95% confidence health and quality of life are presented in the online sup-
intervals. ICBT, Internet-based cognitive behavioral therapy; plementary Table 3.
WLC, waitlist control; Pre, before treatment; Post, after treatment;
6MFU, 6-month follow-up. The ICBT group showed larger improvements from
the pre- to post-treatment assessment compared to the
WLC group on measures of exhaustion, depression,
disturbed sleep and anxiety, with moderate to large be-
Truax [42]. In the present study, clinically significant improve- tween-group effect sizes. ICBT was also superior to the
ment required a participant to make a reliable change on the re- WLC in reducing sickness behavior and somatoform
spective measure (7.12 units on PSS; 1.28 units on SMBQ) as well
as meet criteria for absolute improvement. Absolute improvement
complaints, as well as in increasing self-rated health and
meant that a participant, after treatment, had to have a rating clos- quality of life. There were no significant interaction ef-
er to the normal population than to the clinical population to be fects between ICBT and WLC on functional impair-
classified as a responder. For PSS, the cutoff was defined as a total ment or work ability. However, regarding functional
score of < 31 based on data from Lavoie and Douglas [43]. For impairment, the between group effect size after treat-
SMBQ, the cutoff was established as a total score of <3.9 [44, 45].
Number needed to treat, i.e. the number of patients who need to
ment indicated significantly lower functional impair-
be treated for one additional case in remission, was calculated us- ment in the ICBT group. Between the post-treatment
ing the formula suggested by Kraemer and Kupfer [46]. Patients assessment and 6MFU, all effects were maintained or,
who did not reply to post-treatment and 6MFU measurements as in the case of depression and anxiety, further im-
were considered not improved. proved (F = 5.23–6.78; df = 1, 45–46; p values = 0.012–
0.027).

Results Remission Rates

At the post-treatment assessment, the proportion of
Attrition and Adherence patients in remission (i.e., meeting criteria for clinically
Figure 2 displays a CONSORT diagram of participant significant improvement) on the PSS was 62% in the
flow through the trial. The average number of completed ICBT group and 10% in the WLC group. The number
modules in ICBT was 9.2 out of 12 (SD = 3.2). Nine pa- needed to treat for one case of remission was 1.9 after
tients (18%) did not receive the intended treatment (i.e., treatment. At 6MFU, the remission rate in ICBT in-
completed less than 50% of the treatment modules). The creased to 68%. For the outcome measuring exhaustion
primary reason for non-completion was that the treat- (SMBQ), 48% in the ICBT group and 6% in the WLC were
ment was found too comprehensive (too much text, exer- in remission after treatment. Six months later, this pro-
cises too demanding). The average therapist time per pa- portion increased to 56% in the ICBT group. The differ-
tient over the span of the treatment was 87 min (SD = ence in remission rates between ICBT and WLC after
35.6), i.e. a little over 7 min weekly per patient. treatment was statistically significant for both outcomes
(χ2 = 22.4–29.3; df = 1; p values <0.000).

Internet-Based Cognitive Behavioral Psychother Psychosom 2018;87:296–305 301

Therapy for Chronic Stress DOI: 10.1159/000490742

This content downloaded from

13.234.121.200 on Fri, 15 Sep 2023 13:48:34 +00:00
All use subject to https://about.jstor.org/terms
Table 2. Means, interaction effects and effect sizes (Cohen’s d) on outcome measures for patients with adjustment disorder and exhaus-
tion disorder

Measure Mean (SD) F value2 Effect size (95% CI)

(range1)
pre post 6MFU within group between-group within-group
pre-post post pre-6MFU

PSS (0–56)
ICBT 37.2 (7.1) 24.2 (8.6) 21.9 (7.7) 38.34 (1, 96)*** 1.66 (1.19, 2.10) 1.09 (0.66, 1.51) 2.07 (1.56, 2.54)
WLC 36.4 (7.3) 33.2 (7.9) 0.42 (0.02, 0.82)
SMBQ (1–7)
ICBT 5.3 (0.9) 3.7 (1.3) 3.3 (1.3) 42.04 (1, 95)*** 1.48 (1.02, 1.91) 1.09 (0.66, 1.51) 1.87 (1.37, 2.33)
WLC 5.4 (0.8) 5.0 (1.2)
MADRS–S (0–54)
ICBT 20.4 (8.0) 12.0 (7.6) 9.1 (6.4) 25.89 (1, 96)*** 1.08 (0.65, 1.49) 0.67 (0.26, 1.07) 1.55 (1.09, 1.99)
WLC 18.9 (6.9) 17.4 (8.4) 0.20 (–0.20, 0.60)
ISI (0–28)
ICBT 15.1 (5.4) 6.9 (6.1) 6.5 (5.5) 35.25 (1, 96)*** 1.42 (0.97, 1.85) 1.05 (0.62, 1.47) 1.56 (1.09, 2.00)
WLC 15.3 (6.2) 13.9 (7.2) 0.21 (–0.19, 0.60)
GAD–7 (0–21)
ICBT 11.8 (5.3) 6.2 (4.8) 4.6 (4.0) 26.14 (1, 96)*** 1.11 (0.68, 1.52) 0.57 (0.16, 0.97) 1.52 (1.06, 1.96)
WLC 9.8 (4.2) 9.0 (5.3) 0.17 (–0.23, 0.56)
WHODAS (0–100)
ICBT 25.1 (16.0) 14.9 (14.8) 12.3 (14.7) 3.76 (1, 96) 0.66 (0.25, 1.06) 0.49 (0.08, 0.89) 0.83 (0.41, 1.24)
WLC 28.0 (19.2) 23.2 (18.9) 0.25 (–0.15, 0.65)
WAI (7–49)
ICBT 32.9 (8.8) 36.5 (8.9) 38.4 (8.3) 2.79 (1, 94) 0.40 (–0.79, 0.00) 0.21 (–0.61, 0.19) 0.63 (0.22, 1.04)
WLC 33.1 (7.9) 34.6 (8.4) 0.19 (–0.58, 0.21)

pre, before treatment; post, after treatment; 6MFU, 6-month follow-up; ICBT, Internet-based cognitive behavioral therapy; WLC,
waitlist control; PSS, Perceived Stress Scale; SMBQ, Shirom-Melamed Burnout Questionnaire; MADRS-S, Montgomery-Åsberg Depres-
sion Rating Scale Self-Report; ISI, Insomnia Severity Index; GAD-7, 7-item Generalized Anxiety Disorder Scale; WAI, Work Ability
Index; WHODAS, World Health Organization Disability Assessment Schedule. 1 Range is theoretical, not observed. 2 F values for inter-
action effects between time (pre-post) and group (ICBT vs WLC). *** p < 0.001.

Treatment Credibility, Treatment Satisfaction and
Adverse Events
The mean rating of treatment credibility (scale range
0–50) in ICBT was 37.0 (SD = 7.9) in week 2 and 36.0
(SD = 9.1) in week 8, indicating high treatment credibil-
ity. The mean treatment satisfaction (scale range 8–32) in
the ICBT group was 25.7 (SD = 4.9). Eighty-four percent
of patients reported being satisfied or very satisfied with
the treatment, and 86% would recommend it to a friend.
Nine ICBT patients (18%) reported in total 11 adverse
events. More than half (55%) were related to increased
stress due to treatment workload. All but 2 adverse events
(82%) were classified by patients as having little or no
negative effect on them. The other 2 were classified as
having a very negative effect at the time of occurrence but
did not have any long-term effects.
Receiving Treatment Outside of the Study
Three patients (6%) in the ICBT and 4 (8%) in the
WLC group reported having received other treatment for
stress-related symptoms during the intervention period.
Fisher’s exact test revealed no significant difference be-
tween the groups (p = 0.465).
Subgroup Analyses of AD and ED
Online supplementary Tables 4 and 5 present results
from subgroup analyses, which indicate that the treat-
ment was overall effective in reducing symptoms and im-
proving health-related outcomes in both diagnostic
groups (AD and ED). There were no significant interac-
tion effects of diagnostic group on any outcomes before
to after treatment (F = 0.00–3.33; df = 1, 91–95; p val-
ues = 0.071–0.998) with the exception of self-rated health.
On self-rated health, only patients with ED made statisti-

302 Psychother Psychosom 2018;87:296–305 Lindsäter et al.

DOI: 10.1159/000490742

This content downloaded from

13.234.121.200 on Fri, 15 Sep 2023 13:48:34 +00:00
All use subject to https://about.jstor.org/terms
cally significant improvements compared to patients with
ED in the WLC group (F = 4.67; df = 1, 93; p = 0.033).
There was no difference in completion rate or treatment
satisfaction between the diagnostic groups.
Discussion
The aim of this randomized controlled trial was to in-
vestigate the effect of ICBT on symptoms of chronic stress
in the form of AD and ED. ICBT led to large improve-
ment on the primary outcome of perceived stress and
moderate to large improvement in all secondary symp-
tom domains compared to the control group. Effects were
maintained at the 6MFU. There were no differences be-
tween groups on measures of functional impairment or
work ability. Treatment satisfaction was high while attri-
tion was low and adverse events few. Results indicate that
the employed ICBT protocol can be effective in reducing
stress-related symptoms and increasing quality of life in
a clinical sample suffering from chronic stress.
In the present study, between-group effects on stress
(d = 1.09), depression (d = 0.67) and anxiety (d = 0.57)
were larger than the pooled effects on the same outcomes
found in a meta-analysis of Internet-based interventions
for stress (d = 0.45, 0.34 and 0.32, respectively) [9]. One
reason for this may be that previous studies have used
subclinical or undiagnosed samples, often included
based on a cutoff score on a stress scale. With lower
symptom severity at baseline there is less room for im-
provement, and without clinical assessment samples are
likely heterogeneous and there is a risk of high measure-
ment variance. In the present study, all patients had
moderate to severe pre-treatment symptoms, as is ex-
pected in a clinical sample. Comparing our results to oth-
er studies of ICBT where participants had elevated levels
of stress or were diagnosed with stress-related disorders
[10–12], effects on mental health outcomes were gener-
ally equivalent. The only exception was the effect on dis-
turbed sleep, which was larger in our study (d = 1.05 as
compared to d = 0.34–0.52). It is possible that the early
interventions targeting increased recovery and disturbed
sleep in the present study (weeks 2 and 3 of treatment)
had positive effects on sleep. Previous studies have in-
cluded optional interventions for improved sleep, gener-
ally later in the treatment course [10–12]. Studies are
needed to investigate potential mediators of change in
ICBT for symptoms of chronic stress, particularly the
mediating effect of sleep in improving stress-related out-
comes in CBT.
Internet-Based Cognitive Behavioral
Therapy for Chronic Stress
This content downloaded from
13.234.121.200 on Fri, 15 Sep 2023 13:48:34 +00:00
All use subject to https://about.jstor.org/terms
We find it noteworthy that patients in the ICBT group
gained such overall symptom relief within merely 12
weeks. Indeed, patients with severe symptoms of chronic
stress (ED), of which 40% met criteria for comorbid de-
pression or anxiety disorders, reduced their symptoms of
perceived stress, exhaustion and disturbed sleep with be-
tween-group effect sizes in the range of d = 1.22–1.27 at
the post-treatment assessment. Even though therapists
spent only 7 min per patient and week, effects are compa-
rable to those of face-to-face CBT for stress, as reported
in meta-analyses (d = 0.68–1.16) [5, 6]. It is possible that
the highly structured ICBT format served to reduce ther-
apist drift and thereby compensated for the limited ther-
apist contact.
In spite of the significant symptom reduction, partici-
pants in ICBT did not reduce their functional impairment
or improve work ability relative to the WLC. Given that
the observed improvement on functional impairment
was moderate in size (d = 0.49, 95% CI = 0.08–0.89) with
a p value on the border of being significant (p = 0.055), we
however believe that it is premature to conclude that
ICBT does not have an effect on this outcome. Larger
sample sizes and longer follow-up periods are needed to
establish potential effects of ICBT on functional impair-
ment. Regarding work ability, results from the present
study support previous findings that decreased symp-
toms are not necessarily related to perceived work ability
or return to work after sick leave [4, 23]. The lack of treat-
ment effect could partly be understood by the fact that the
treatment’s primary focus was on mental health rather
than on work-related difficulties. Another explanation
could be that patients in the present study in fact reported
moderate work ability before treatment, which possibly
limited the room for improvement within this area. Fur-
ther research is needed to understand the complex rela-
tionship between symptom reduction, functional impair-
ment and work ability.
Important strengths of this study were the randomized
design, validated treatment outcomes, low attrition rates
and high treatment acceptability. A limitation was the use
of a WLC crossing-over to active treatment, which did
not allow for between-group comparisons at the 6MFU
or enable control over potential confounders between the
post-treatment and 6MFU assessments (e.g., seasonal ef-
fects). It is possible that factors such as treatment settings
and patient’s expectations in ICBT and WLC affected
outcomes [47], and conclusions must be drawn with cau-
tion. However, using a WLC provides ethical advantages
(treatment guaranteed) as well as a control for regression
towards the mean and spontaneous remission [48].
Psychother Psychosom 2018;87:296–305 303
DOI: 10.1159/000490742
 Another limitation pertains to the method of assess-
ment. Due to the nationwide recruitment, conventional
face-to-face assessment was not possible. While taking
advantage of a central strength of Internet-based treat-
ments, i.e. that there is no restriction in terms of geo-
graphic reach, conducting telephone assessments may
lead to information being missed. Further, use of the clin-
ical diagnoses AD and ED carries some difficulties. As
mentioned in the Introduction, AD is a debated diagnosis
[e.g., 14, 16] and ED has limited international recogni-
tion. Indeed, it is challenging to fit the complexity of
chronic stress, which comprises both biological, psycho-
logical and social factors, into customary taxonomies of
medical disease [13, 18]. In the face of this complexity and
the lack of previous research, we consider it a strength of
the study to use clinical diagnoses defined by DSM-5 and
ICD-10 as a means of operationalizing clinically signifi-
cant symptoms associated with chronic stress. This aids
in replication of the study and means that results are like-
ly to generalize to those most in need of psychological
treatment.
A final limitation concerns generalizability of findings.
We used a voluntary sample consisting of highly educated
85% female participants, and only 2 therapists conducted
treatment. Although therapist effects in ICBT are indi-
cated to be minimal [e.g., 49], future studies might in-
crease the number of involved therapist and collect data
on treatment fidelity. Further, future studies should rep-
licate the present study using different recruitment strat-
egies (e.g., consecutive recruitment in primary care set-
tings) and active control conditions.
In summary, the present study supports previous find-
ings that ICBT can be efficacious in reducing symptoms
of stress. It adds to the knowledge base by testing the ef-
ficacy of ICBT in a clinically assessed and diagnosed sam-
ple suffering from chronic stress. Considering the mini-
mal therapist time required and the relative simplicity of
the treatment, ICBT could be a highly cost-effective treat-
ment with the potential to substantially increase treat-
ment accessibility to those most in need.
Acknowledgments
This research was funded by the Doctoral School in Health
Care Sciences, Karolinska Institutet, Stockholm Stress Center and
Stockholm County Council. These are public institutions neither
of which had any role in the design, execution or publication of the
study.
Disclosure Statement
All authors report that they have no competing interests or any
financial interest in the material presented.

References
1 McEwen BS: Physiology and neurobiology of
stress and adaptation: central role of the brain.
Physiol Rev 2007;87:873–904.
2 Grossi G, Perski A, Osika W, Savic I: Stress-
related exhaustion disorder – clinical mani-
festation of burnout? A review of assessment
methods, sleep impairments, cognitive dis-
turbances, and neuro-biological and physio-
logical changes in clinical burnout. Scand J
Psychol 2015;56:626–636.
3 Kalia M: Assessing the economic impact of
stress – the modern day hidden epidemic.
Metabolism 2002;51:49–53.
4 Henderson M, Harvey SB, Overland S, Mykle-
tun A, Hotopf M: Work and common psychi-
atric disorders. J R Soc Med 2011; 104: 198–
207.
5 Van der Klink JJ, Blonk RW, Schene AH, van
Dijk FJ: The benefits of interventions for
work-related stress. Am J Public Health 2001;
91:270–276.
6 Richardson KM, Rothstein HR: Effects of oc-
cupational stress management intervention
programs: a meta-analysis. J Occup Health
Psychol 2008;13:69–93.
7 Cavanagh K: Geographic inequity in the
availability of cognitive behavioural therapy
in England and Wales: a 10-year update. Be-
hav Cogn Psychother 2014;42:497–501.
8 Hedman E, Ljotsson B, Lindefors N: Cogni-
tive behavior therapy via the Internet: a sys-
tematic review of applications, clinical effica-
cy and cost-effectiveness. Expert Rev Phar-
macoecon Outcomes Res 2012;12:745–764.
9 Heber E, Ebert DD, Lehr D, Cuijpers P, Berk-
ing M, Nobis S, Riper H: The benefit of web-
and computer-based interventions for stress:
a systematic review and meta-analysis. J Med
Internet Res 2017;19:e32.
10 Heber E, Lehr D, Ebert DD, Berking M, Riper
H: Web-based and mobile stress management
intervention for employees: a randomized
controlled trial. J Med Internet Res 2016; 18:
e21.
11 Ebert DD, Lehr D, Heber E, Riper H, Cuijpers
P, Berking M: Internet- and mobile-based
stress management for employees with adher-
ence-focused guidance: efficacy and mecha-
nism of change. Scand J Work Environ Health
2016;42:382–394.
12 Persson Asplund R: Internet-based stress
management for distressed managers: results
from a randomised controlled trial. JMIR Res
Protoc 2018;75:105–113.
13 Henningsen P, Zipfel S, Sattel H, Creed F:
Management of functional somatic syn-
dromes and bodily distress. Psychother Psy-
chosom 2018;87:12–31.
14 Carta MG, Balestrieri M, Murru A, Hardoy
MC: Adjustment disorder: epidemiology, di-
agnosis and treatment. Clin Pract Epidemiol
Ment Health 2009;5:15.
15 American Psychiatric Association: Diagnos-
tic and Statistical Manual of Mental Disor-
ders: DSM-5. Arlington, American Psychiat-
ric Association, 2013.

304 Psychother Psychosom 2018;87:296–305 Lindsäter et al.

DOI: 10.1159/000490742

This content downloaded from

13.234.121.200 on Fri, 15 Sep 2023 13:48:34 +00:00
All use subject to https://about.jstor.org/terms
16 Casey P: Adjustment disorder: new develop-
ments. Curr Psychiatry Rep 2014;16:451.
17 World Health Organization: The ICD-10
Classification of Mental and Behavioural Dis-
orders: Clinical Descriptions and Diagnostic
Guidelines. Geneva, World Health Organiza-
tion, 1992.
18 Fava GA, Cosci F, Sonino N: Current psycho-
somatic practice. Psychother Psychosom 2017;
86:13–30.
19 Schaufeli W, Enzmann D: The Burnout Com-
panion to Study and Practice: A Critical Anal-
ysis. Boca Raton, CRC Press, 1998.
20 Sheehan DV, Lecrubier Y, Sheehan KH,
Amorim P, Janavs J, Weiller E, Hergueta T,
Baker R, Dunbar GC: The Mini-International
Neuropsychiatric Interview (M.I.N.I.): the
development and validation of a structured
diagnostic psychiatric interview for DSM-IV
and ICD-10. J Clin Psychiatry 1998; 59(suppl
20):22–33.
21 Rohde P, Lewinsohn PM, Seeley JR: Compa-
rability of telephone and face-to-face inter-
views in assessing axis I and II disorders. Am
J Psychiatry 1997;154:1593–1598.
22 Salomonsson S, Santoft F, Lindsater E, Ejeby
K, Ljotsson B, Ost LG, Ingvar M, Lekander M,
Hedman-Lagerlof E: Stepped care in primary
care – guided self-help and face-to-face cogni-
tive behavioural therapy for common mental
disorders: a randomized controlled trial. Psy-
chol Med 2018;48:1644–1654.
23 Salomonsson S, Santoft F, Lindsater E, Ejeby
K, Ljotsson B, Ost LG, Ingvar M, Lekander M,
Hedman-Lagerlof E: Cognitive-behavioural
therapy and return-to-work intervention for
patients on sick leave due to common mental
disorders: a randomised controlled trial. Oc-
cup Environ Med 2017;74:905–912.
24 Morin CM, Benca R: Chronic insomnia. Lan-
cet 2012;379:1129–1141.
25 Lejuez CW, Hopko DR, Acierno R, Daughters
SB, Pagoto SL: Ten year revision of the brief
behavioral activation treatment for depres-
sion: revised treatment manual. Behav Modif
2011;35:111–161.
26 Abramowitz JS, Deacon BJ, Whiteside SP: Ex-
posure Therapy for Anxiety: Principles and
Practice. New York, Guilford Press, 2012.
27 Cohen S, Kamarck T, Mermelstein R: A glob-
al measure of perceived stress. J Health Soc
Behav 1983;24:385–396.
Internet-Based Cognitive Behavioral
Therapy for Chronic Stress
13.234.121.200 on Fri, 15 Sep 2023 13:48:34 +00:00
28 Eklund M, Backstrom M, Tuvesson H: Psy-
chometric properties and factor structure of
the Swedish version of the Perceived Stress
Scale. Nord J Psychiatry 2014;68:494–499.
29 Lundgren-Nilsson A, Jonsdottir IH, Pallant J,
Ahlborg G Jr: Internal construct validity of
the Shirom-Melamed Burnout Questionnaire
(SMBQ). BMC Public Health 2012;12:1.
30 Svanborg P, Åsberg M: A new self-rating scale
for depression and anxiety states based on the
Comprehensive Psychopathological Rating
Scale. Acta Psychiatr Scand 1994;89:21–28.
31 Morin CM, Savard J, Ouellet MC, Daley M:
Insomnia. Wiley Online Library, 1993.
32 Spitzer RL, Kroenke K, Williams JB, Lowe B:
A brief measure for assessing generalized anx-
iety disorder: the GAD-7. Arch Intern Med
2006;166:1092–1097.
33 Andreasson AN, Wicksell R, Karshikoff B,
Lodin K, Axelsson J, Höglund CO, Lekander
M: Development and preliminary validation
of the Sickness Questionnaire (SicknessQ).
Brain Behav Immun 2013;32:e14.
34 Eriksson I, Unden AL, Elofsson S: Self-rated
health. Comparisons between three different
measures. Results from a population study.
Int J Epidemiol 2001;30:326–333.
35 Van Ravesteijn H, Wittkampf K, Lucassen P,
van de Lisdonk E, van den Hoogen H, van
Weert H, Huijser J, Schene A, van Weel C,
Speckens A: Detecting somatoform disorders
in primary care with the PHQ-15. Ann Fam
Med 2009;7:232–238.
36 Lindner P, Frykheden O, Forsstrom D, An-
dersson E, Ljotsson B, Hedman E, Andersson
G, Carlbring P: The Brunnsviken Brief Qual-
ity of Life Scale (BBQ): development and Psy-
chometric Evaluation. Cogn Behav Ther 2016;
45:182–195.
37 Axelsson E, Lindsäter E, Ljótsson B, Anders-
son E, Hedman-Lagerlöf E: The 12-item Self-
Report World Health Organization Disability
Assessment Schedule (WHODAS) 2.0 ad-
ministered via the Internet to individuals with
anxiety and stress disorders: a psychometric
investigation based on data from two clinical
trials. JMIR Ment Health 2017;4:e58.
Psychother Psychosom 2018;87:296–305
DOI: 10.1159/000490742
This content downloaded from
All use subject to https://about.jstor.org/terms
38 Tuomi K, Ilmarinen J, Jahkola A, Katajarinne
L, Tulkki A: Work ability index. Helsinki,
Finnish Institute of Occupational Health,
1998.
39 Borkovec TD, Nau SD: Credibility of ana-
logue therapy rationales. J Behav Ther Exp
Psychiatry 1972;3:257–260.
40 Larsen DL, Attkisson CC, Hargreaves WA,
Nguyen TD: Assessment of client/patient sat-
isfaction: development of a general scale. Eval
Program Plann 1979;2:197–207.
41 Gueorguieva R, Krystal JH: Move over ANO-
VA: progress in analyzing repeated-measures
data and its reflection in papers published in
the Archives of General Psychiatry. Arch Gen
Psychiatry 2004;61:310–317.
42 Jacobson NS, Truax P: Clinical significance: a
statistical approach to defining meaningful
change in psychotherapy research. J Consult
Clin Psychol 1991;59:12–19.
43 Lavoie JAA, Douglas KS: The Perceived Stress
Scale: evaluating configural, metric and scalar
invariance across mental health status and
gender. J Psychopathol Behav Assess 2011;34:
48–57.
44 Jonsdottir IH, Hagg DA, Glise K, Ekman R:
Monocyte chemotactic protein-1 (MCP-1)
and growth factors called into question as
markers of prolonged psychosocial stress.
PLoS One 2009;4:e7659.
45 Glise K, Ahlborg G Jr, Jonsdottir IH: Course
of mental symptoms in patients with stress-
related exhaustion: does sex or age make a dif-
ference? BMC Psychiatry 2012;12:18.
46 Kraemer HC, Kupfer DJ: Size of treatment ef-
fects and their importance to clinical research
and practice. Biol Psychiatry 2006; 59: 990–
996.
47 Fava GA, Guidi J, Rafanelli C, Rickels K: The
clinical inadequacy of the placebo model and
the development of an alternative conceptual
framework. Psychother Psychosom 2017; 86:
332–340.
48 Hart T, Fann JR, Novack TA: The dilemma of
the control condition in experience-based
cognitive and behavioural treatment research.
Neuropsychol Rehabil 2008;18:1–21.
49 Andersson G: The promise and pitfalls of the
Internet for cognitive behavioral therapy.
BMC Med 2010;8:82.
305