Insomnia is a prevalent condition (affecting 10% of adults, 10-20% in primary care samples) whose key features are prolonged poor sleep, which affects daytime function. It is associated with numerous physical and mental health problems and has a substantial impact on wellbeing. Struggles with sleep are often one of the top problems patients report when they see mental health services.

Options for treating insomnia include medications and psychological interventions. Medications can be effective, but are often associated with side effects like daytime over-sedation and tolerance and can be addictive. Psychological interventions in current guidelines focus on an adapted form of cognitive behavioural therapy for insomnia (CBTi), which has been adapted and computerised for app-based use.

The core features of CBTi are work on negative cognitions associated with poor sleep (e.g. beliefs that it will not be possible to get to sleep), alongside behavioural interventions (sometimes known as stimulus control). These include having the sleeping environment be as quiet and dark as possible, only using the bedroom for sleep and sex, and avoiding wake-promoting or sleep-disrupting stimuli.

Although CBTi is effective, it is not always practical or possible for patients to use behavioural techniques. For example, due to housing circumstances it may not be possible to ensure the bedroom is reserved for sleep, or achieve full quiet or darkness.

Given the limitations of medications for insomnia, clinicians working in many settings need alternative, evidence-based psychological interventions for insomnia.

This is the gap that Rafihi-Ferreira and colleagues set out to fill, by performing a randomised controlled trial that compares CBTi with both a waiting list control and Acceptance and Commitment Therapy (ACT) for treating insomnia in a group of adult patients. ACT is a third wave therapy that incorporates elements of mindfulness and works on improving psychological flexibility. It has previously been shown to be effective in treating insomnia (Salari et al 2020).

Methods

This randomised controlled trial compared ACT to both CBTi and a control condition. The investigators randomised 227 adults to either to 6 weekly group-based online ACT sessions (n = 76), CBTi (n = 76), or a waitlist control condition (n = 75). Participants were recruited using advertisements on their University social media from around the São Paulo area of Brazil. Inclusion criteria were: age 18-59, self-reporting chronic symptoms of insomnia with difficulty in either delayed sleep onset (≥30 minutes), increased waking after sleep onset (≥30 minutes), or early waking (waking ≤ 6.5 hours after sleep onset), with duration > 3 nights per week, > 3 months duration, causing daytime difficulties.

Participants with an unstable or progressive physical or mental health condition were excluded, including anyone with a lifetime diagnosis of a psychotic or bipolar affective disorder, as was anyone with a history of sleep disorders (e.g. sleep apnoea), substance misuse, and those who were unable to engage (e.g. due to a lack of a computer or working shifts meaning they could not make the sessions).

The primary outcome measured by the investigators was the Insomnia Severity Index (ISI), and also sleep diaries, psychiatric questionnaires and a measure of adherence and satisfaction with the intervention. Participants completed questionnaires at baseline, 2 weeks after the end of their intervention and at a 6-month review.

Results

Of 227 participants randomised (76 to ACT, 76 to CBTi and 75 to waiting list), 199 completed to the end of treatment (66 for ACT, 61 for CBTi and 72 for waiting list), and 191 were followed up at 6 months (64 for ACT, 58 for CBTi and 69 for waiting list). Participants were mostly female (76%), with an average age of 40.6 years, 73% white ethnicity and were well educated, with 81.1% having a university degree (whereas around 12% of the population of Brazil have tertiary education). There were no differences in participant characteristics between the three groups.

The authors found that both CBTi and ACT were more effective in reducing the ISI than a waiting list control, both immediately post-treatment and after 6 months. ACT was found to reduce ISI by 7.91 points after treatment, with a standard deviation (SD) of 1.43 points; CBTi reduced ISI by 10.02 points (SD 1.74), whereas in the waiting list condition ISI reduced by 3.53 points (SD 0.67). At 6 months, there were no significant further changes in ISI score in any treatment group.

In direct comparison, CBTi was more effective than ACT at both the post-treatment and 6 months time-point, but the difference between treatments was small; post-treatment the difference between groups had a Cohen’s d value of 0.26; at 6 months the difference had a d value of 0.21. Both of these effect sizes are considered small differences. By contrast, the effective size of ACT compared to waiting list was -0.57 post-treatment and -0.61 at 6 months, which is considered a moderate effect size.

The authors also calculated other measures including “remission”: an ISI score >8. After treatment with ACT, 19% of participants achieved remission; for CBTi this was 32%, but only 1% of waiting list participants. In direct comparison, CBTi was more effective at inducing remission than ACT post-treatment, but there was no difference between therapies at the 6-month time-point. Both therapies were consistently more effective in promoting remission than waiting list. The authors argue as a result, that ACT has a delayed beneficial effect, as ACT was no longer statistically inferior to CBTi at the 6-month time point.

The authors found similar patterns of results in sleep diary measures e.g., sleep onset latency, waking after sleep onset, sleep efficiency and sleep satisfaction: both ACT and CBTi were more effective than waiting list. Both therapies were also effective in reducing measures of depression and improving psychological flexibility and acceptance. Both CBTi and ACT scored highly, and similarly, for measures of participant understanding, motivation, satisfaction, and for recommending to others. Participants reported being more likely to have completely followed the intervention program recommendations for CBTi compared to ACT (57% for CBTi, 38% for ACT).

Conclusions

The authors concluded:

Our results add to the substantial existing evidence that CBT is an effective treatment for insomnia, even when performed in a group or in an online format. An additional contribution of this study is that ACT used singly, that is, without stimulus control and sleep restriction techniques, is also effective for insomnia… Both treatment groups improved significantly, with large effect sizes for the primary outcome, insomnia severity. The results were maintained at follow-up with large effect sizes… A comparison between therapies showed that CBT-I showed better results than ACT-I in reducing ISI scores but with small effect sizes.

Strengths and limitations

Strengths

This study benefits from clearly stated objectives, is well written and the authors appear to have used robust methods in the design and analysis. They make reasonable and measured conclusions that are supported by the results presented. The authors provide clear information about all participants and account for their flow through the study. They used an appropriate randomisation strategy, stratifying by insomnia severity index, and groups were well balanced. They used an intention-to-treat analysis strategy, which analyses all participants in the groups they were assigned to, and used an analysis approach (mixed models) which allowed them to account for missing data.

The study included clinically useful ways of measuring the impact of the treatment, including both quantitative and qualitative scales. The main findings, that ACT can reduce ISI, had a “moderate” effect size, with reasonably small standard deviations (suggesting consistency effects across participants), indicating potentially meaningful clinical applicability.

A key strength of this study is the long follow-up period (6 months) which allows readers to see that treatment effects appear to be sustained. The authors also managed to follow-up a good percentage of participants (76% of CBTi and 84% of ACT groups) to 6 months, further strengthening the validity of their longitudinal findings.

Providing a group-based therapy online is also practically useful as many mental health services provide interventions through electronic means, and given resource constraints a 6 session intervention is likely to be more achievable than longer therapies. Therefore, the format used in the study aligns well with real life practice.

Limitations

This study has a number of limitations to consider when interpreting its conclusions. As with all psychological intervention studies, both the participants and therapists were inevitably not blinded to their treatment group and the waiting list group had fewer clinical contacts than the intervention groups. This could induce stronger “placebo” responses in participants who know they are in the active therapy groups, compared to participants who know they are on a waiting list.

The inclusion of a waiting list control is helpful in providing evidence that ACT is better than doing nothing for patients. However, given we already have an evidence base for the effectiveness of CBTi, and the core purpose of this study being the comparison between CBTi and ACT, I wonder if it would have been better to make the study a direct comparison of CBTi and ACT with the full study sample. This would have increased the statistical power of the study to determine if ACT or CBTi significantly differed in effectiveness. This would have also meant the participants in the waiting list group would have been able to access some therapy for insomnia.

One of the exclusion criteria for the study was if participants had a major mental health condition. While this is understandable for the purposes of this specific study question, it also limits the application of the study findings to mental health services, who will be treating clients who have been given such diagnoses.

Implications for practice

This study provides a useful addition to our understanding of psychological interventions for insomnia because it shows another form of talking therapy – ACT – is a viable option for treating insomnia, providing an alternative to CBTi. As the authors suggest, some patients do very much struggle with the behavioural techniques in CBTi, so having an alternative psychological option is theoretically very useful. The authors demonstrating the effectiveness of treatment delivered by internet in a group setting is also very useful as it suggests a less resource-intense intervention can be effective.

Pragmatically, my caution about the usefulness of this finding, however, is that I am not clear whether there is an additional pool of ACT therapists available to take on the task of providing this therapy, especially in healthcare systems with limited resources and long waiting lists.

It may perhaps be more practical for existing therapists or mental health professionals, who frequently work with patients experiencing insomnia alongside other mental health concerns (e.g. mental health nurses, social workers, occupational therapists or psychiatrists in community mental health teams) to incorporate some of the elements of ACT into the repertoire of interventions they provide. Being able to offer alternative options to patients who struggle with the behavioural techniques in CBTi might help improve options for insomnia treatment.

This study also contributes to the literature that suggests treatment for insomnia, including CBTi, and now ACT, can improve general psychological measures like flexibility, and improve mood symptoms. This should encourage general mental health practitioners to enquire about sleep and offer interventions about sleep for their clients. This might improve therapeutic engagement and hope that things can get better.

Statement of interests

I have no personal or professional link to this study or its authors.

Links

Primary paper

El Rafihi-Ferreira R, Hasan R, Toscanini AC, Linares IMP, Suzuki Borges D, Brasil IP, Carmo M, Lotufo Neto F, Morin C. (2024) Acceptance and commitment therapy versus cognitive behavioral therapy for insomnia: A randomized controlled trial. J Consult Clin Psychol. 2024        10.1037/ccp0000881     

Other references

Salari N, Khazaie H, Hosseinian-Far A, Khaledi-Paveh B, Ghasemi H, Mohammadi M, Shohaimi S. The effect of acceptance and commitment therapy on insomnia and sleep quality: A systematic review. BMC Neurol. 2020 Aug 13;20(1):300. doi: 10.1186/s12883-020-01883-1

Photo credits

• Photo by Jay Wennington on Unsplash
• Photo by Wilhelm Gunkel on Unsplash