Online CBT Effective for Insomnia, Psychosis

Pauline Anderson

September 13, 2017

An Internet-based cognitive-behavioral therapy (CBT) intervention aimed at improving sleep not only significantly reduces insomnia but also cuts symptoms of paranoia and hallucinations, results of a large, randomized trial show. The intervention also appears to quell anxiety and depressive symptoms.

These results imply that sleep problems play an important causal role in mental health problems and that treating sleep may also help some psychotic conditions.

"The really interesting implication of this work is that if you treat sleep problems ― and we have some very good treatments ― a whole cascade of other benefits for psychiatric health follow," lead study author Daniel Freeman, PhD, professor of clinical psychology, Oxford University, United Kingdom, told Medscape Medical News.

Disrupted sleep has been "overlooked" as a factor that causes or provokes mental health problems. In mental health services, "there is far too little sleep training for staff, far too little formal assessment of sleep problems, and far too little evidence-based treatments given," he added.

The study was published online September 6 in Lancet Psychiatry.

"Cartoon" Therapist

Insomnia affects about 5% to 10% of the population. The rate of insomnia is twice as high in women than in men.

Clinical guidelines recommend CBT as first-line treatment. In addition to face-to-face CBT, digital forms, which do not require a therapist to be present, have also shown efficacy.

The single-blind Oxford Access for Students Improving Sleep (OASIS) trial included adult students attending a UK university who suffered from insomnia, defined as having a score of 16 or lower on the Sleep Condition Indicator (SCI).

Of those screened, 1864 were randomly assigned to receive usual care (ie, whatever the participants were already receiving, which typically meant no care), and 1891 were assigned to the intervention group.

The CBT intervention, known as Sleepio, includes six sessions, each lasting about 20 minutes. It involves behavioral, cognitive, and educational components.

Behavioral techniques include sleep restrictions, stimulus control, and relaxation. The cognitive component includes such techniques such as imagery, mindfulness, and belief restructuring, for example, targeting unrealistic expectations about sleep. The educational portion includes information about sleep hygiene and processes involved in sleep.

The program is interactive, and content is presented by an animated or "cartoon" therapist.

Study participants were able to access the interactive sessions as well as sleep diaries, relaxation audios, and other tools using the Web browser of their smartphone. Access was "incredibly easy," said Dr Freeman.

The primary outcomes were insomnia, paranoia, and hallucinations.

Researchers measured insomnia using the SCI-8, which assesses sleep and its impact on daytime functioning during the past week. They assessed paranoia during the past fortnight with the Paranoid Thought Scales, part B, which assesses persecutory ideation. And they measured hallucinations with the Specific Psychotic Experiences Questionnaire-Hallucinations subscale.

Assessments took place at weeks 0 (baseline), 3, 10 (end of therapy), and 22. A mediation analysis was carried out at week 3 to see whether changes in sleep at that time preceded changes in paranoia and hallucinations.

The sample was predominantly female, and two thirds were British.

High Dropout Rate

At 50%, the dropout rate was relatively high and was greater in the treatment group. In part, this may be due to the fact the study was conducted completely online. Study participants may have been less likely to drop out if they had had to check in with a live person, said Dr Freeman.

Treatment uptake was relatively low. Although 69% of the participants logged in for at least one treatment session, progressively smaller percentages logged into additional sessions.

The intervention resulted in significant changes compared to the control at week 10 in insomnia (adjusted difference 4.78; 95% confidence interval [CI], 4.29 - 5.26), paranoia (adjusted difference, -2.22; 95% CI, -2.98 to -1.45), and hallucinations (adjusted difference, -1.58; 95% CI, -1.98 to -1.18; all P < .0001).

After treatment, 62% in the treatment group and 29% in the control group scored outside the clinical cutoff point for insomnia that was used for trial entry.

The mediation analysis provided further support for the hypothesis that disrupted sleep has a causal role in psychotic experiences. This analysis showed that improvements in sleep at 3 weeks accounted for almost 60% of the change in paranoia and for almost 40% of the change in hallucinations after treatment.

Reductions in paranoia and hallucinations affected sleep "hardly at all," said Dr Freeman.

The intervention also resulted in improvements in depression and in anxiety, prodromal symptoms, nightmares, psychological well-being, and functioning. All of these improvements were maintained over time.

Contact with mental health services did not differ between groups. However, Dr Freeman noted that it probably would have taken a longer follow-up period to detect such changes.

The sleep treatment led to a small sustained increase in symptoms of mania. Dr Freeman noted that the assessment of mania included such things as "self-confidence" and "needing less sleep" and that results may "simply reflect generally improved mood."

More Easily Accessible

In an accompanying editorial coauthored by Tea Lallukka, PhD, Finnish Institute of Occupational Health and the Department of Public Health, University of Helsinki, noted that online therapy is an effective strategy for reaching large numbers of people.

"As opposed to traditional face-to-face interventions, this web-based treatment is more easily accessible to participants as treatment could be accessed at any convenient time or place."

Although CBT is recommended as a first-line treatment for the management of chronic insomnia, its use through live therapists is limited because of high cost and low availability, said Dr Lallukka.

Although the dropout rate in the study was high, the authors provided "compelling analyses" that assessed the potential effects of dropout and selectiveness, and, "importantly," outcomes were not associated with "missingness," she said.

The study authors argue that because the treatment and the observed effects were similar to those in previous studies in general adult populations, the findings might be transferable beyond student samples, said Dr Lallukka.

"Still, a need clearly exists for future studies to address some of the shortcomings that exist in the current published work, including recruiting other demographic groups, and using longer follow-up times, in order to examine whether treatment gains and relapse prevention are maintained."

A Novel Focus

Commenting on the study for Medscape Medical News, Matthew R. Ebben, PhD, associate professor of psychology in clinical neurology, Center for Sleep Medicine, Weill Medical College of Cornell University, New York City, said, "There's no question that this is a good study."

Numerous studies have shown that CTB is effective for insomnia, that treating insomnia reduces depression and anxiety, and that treating depression and anxiety improves sleep, said Dr Ebben.

What makes this study "novel" is the focus on psychosis or hallucinations.

He found it unusual that so many patients in the study (38%) started out at ultra–high risk for psychosis, with a score of 6 or greater on the 16-item version of the Prodromal Questionnaire. (That questionnaire served as a secondary outcome measure for psychotic experiences.)

"That seems to be surprisingly high," given that "the prevalence of psychotic disease tends to be in the 3% to 4% range overall."

However, he noted that the Prodromal !uestionnaire is more a measure of having a "propensity" to psychosis.

Dr Ebben also pointed out that, despite the large size of the study, effect sizes for the outcomes were relatively small. "And the effect on paranoia and hallucinations was much, much smaller than the effect on insomnia."

Dr Freeman reports receiving grants from the Wellcome Trust, nonfinancial support (provision of the sleep treatment for the OASIS trial) from Sleepio, grants from the UK National Institute of Health Research, and personal fees from Oxford Virtual Reality, a University of Oxford spinoff company. Dr Ebben and the editorialists report no relevant financial relationships.

Lancet Psychiatry. Published online September 6, 2017. Full text, Editorial

For more Medscape Psychiatry news, join us on Facebook and Twitter


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.