News · Sleep
Online sleep therapy worked for autistic adults, a first trial
A SLEEP randomized trial of 163 autistic adults found a 5-week internet CBT course cut insomnia severity substantially, with gains holding at six months.
Based on a peer-reviewed randomized trial in SLEEP
- Researchers at Vrije Universiteit Amsterdam ran the first randomized trial of autism-adapted internet CBT for insomnia, publishing in the journal SLEEP.
- 163 autistic adults with insomnia were randomized to a 5-week guided self-help online program (82) or a waitlist receiving sleep education (81).
- The program was co-created with seven autistic adults who had sleeping problems, rather than simply repurposed.
- Insomnia severity fell substantially in the intervention group, with a large effect size (d = 1.02; 95% CI, 0.65-1.40), and gains held at 6-month follow-up.
- 37.8% of the intervention group reached clinically meaningful improvement versus 11.1% of controls.
- Quality of life and pre-sleep arousal also improved. Anxiety, depression and daily functioning did not.
- Only half completed all five modules, and roughly six in ten did not reach clinically meaningful improvement.
Roughly six in ten autistic adults have insomnia, and almost none of the treatment research has been designed with them in mind. A trial in the journal SLEEP reports the first attempt to fix that, and it worked.
An estimated 60% of autistic adults suffer from insomnia symptoms, which significantly impairs daily functioning and wellbeing. A five-week online course, built with autistic adults rather than for them, cut that substantially.
Adapted, not repurposed
The distinction matters. Standard cognitive behavioral therapy for insomnia is the recommended first-line treatment for the general population, and simply handing it to a different group is not a trial worth running.
Instead, in co-creation with seven autistic adults with sleeping problems, i-Sleep Autism was adapted from i-Sleep, a pre-existing non-autism specific intervention.
The authors describe the result as the first randomized controlled trial to investigate autism-adapted internet therapy for autistic adults suffering from insomnia.
The trial
A total of 163 adults with autism and insomnia was randomly allocated to i-Sleep Autism, 82 people, or waitlist control, 81 people, receiving psychoeducation and sleep hygiene. Note that the comparison group was not left with nothing; they got the standard advice.
Participants were assessed at baseline, mid-intervention at three weeks, post-test at six weeks, and follow-up at six months, on insomnia severity as the primary outcome plus pre-sleep arousal, anxiety, depression, daily functioning and quality of life.
What improved
The analysis revealed a significant reduction in insomnia severity at mid-intervention and post-test, and in cognitive and somatic pre-sleep arousal at post-test, comparing intervention to control. The effect size was large.
The clinically meaningful threshold is the number that translates. At post-test, 37.8% of the intervention group versus 11.1% of the control group showed clinically meaningful improvement of insomnia.
Crucially, it lasted: effects on insomnia severity were maintained at 6-month follow up. The intervention group also showed significantly higher quality of life at post-test.
What did not improve
The paper is refreshingly plain about the ceiling. No significant effects were found for anxiety, depression, or daily functioning.
Better sleep is often sold as a lever on mood and function. In this trial, over this timescale, it was not. The gain was specific to sleep and to quality of life, and stopping there is more useful than implying otherwise.
The people it missed
Roughly six in ten in the intervention group did not reach the clinically meaningful threshold. Engagement was partial too: in the intervention group, 56.1% completed four, and 50% all five modules.
Half the participants finished the course. That is not unusual for guided self-help, and it sets a realistic expectation for anyone considering this route.
The authors close on exactly that point, concluding that i-Sleep Autism effectively and sustainably improves insomnia, and that additional strategies are needed to support dropouts and those who did not benefit.
For a group whose sleep problems have gone largely unstudied, a five-week online course that works for a third of people and holds for six months is a reasonable place to start.
People also ask
Why do autistic adults need a different sleep program?
Insomnia is far more common in this group, with an estimated 60% of autistic adults affected. Standard CBT for insomnia assumes things that may not hold, including comfort with ambiguous instructions, tolerance of routine disruption and sensory environments that suit the average sleeper. This program was adapted in co-creation with seven autistic adults with sleeping problems, which is why the trial is testing a genuinely different intervention rather than the same one in a new population.
How large is the effect?
Large by the standards of psychological interventions. The reduction in insomnia severity carried an effect size of 1.02, where 0.8 is conventionally considered large. More concretely, 37.8% of the intervention group showed clinically meaningful improvement against 11.1% of controls, and the benefit was still present six months later, which is where many sleep interventions fade.
Did it help with anxiety or depression too?
No, and the paper reports that plainly: no significant effects were found for anxiety, depression, or daily functioning. That is worth knowing, because sleep interventions are often promoted as improving mental health generally. Here, better sleep did not translate into measurably better mood over the study period, though quality of life did improve.
What about the people it did not help?
They are the majority, and the authors say so. Roughly six in ten in the intervention group did not reach the clinically meaningful threshold, and only half finished all five modules. The paper closes by noting that additional strategies are needed to support dropouts and those who did not benefit. A self-guided online course suits some people and not others.
Is this available to use?
The program, i-Sleep Autism, is a research intervention rather than a commercial product, and this is one trial. It also used guided self-help, meaning a supporter checked in, not a fully automated app. Anyone with persistent insomnia should raise it with a doctor; CBT for insomnia is the recommended first-line treatment generally, and this trial suggests an adapted version can work for autistic adults specifically.