BackgroundCognitive behavioral therapy is treatment of choice for insomnia, but availability is scarce. Self-help can increase availability at low cost, but evidence for its efficacy is limited, especially for the typical insomnia patient with co-morbid problems. We hypothesized that a cognitive behaviorally based self-help book is effective to treat insomnia in individuals, also with co-morbid problems, and that the effect is enhanced by adding brief therapist telephone support.MethodsVolunteer sample; 133 media-recruited adults with insomnia. History of sleep difficulties (mean [SD]) 11.8 [12.0] years. 92.5% had co-morbid problems (e.g. allergy, pain, and depression). Parallel randomized (block-randomization, n ≥ 21) controlled "open label" trial; three groups-bibliotherapy with (n = 44) and without (n = 45) therapist support, and waiting list control (n = 44). Assessments before and after treatment, and at three-month follow-up. Intervention was six weeks of bibliotherapeutic self-help, with established cognitive behavioral methods including sleep restriction, stimulus control, and cognitive restructuring. Therapist support was a 15-minute structured telephone call scheduled weekly. Main outcome measures were sleep diary data, and the Insomnia Severity Index.ResultsIntention-to-treat analyses of 133 participants showed significant improvements in both self-help groups from pre to post treatment compared to waiting list. For example, treatment with and without support gave shorter sleep onset latency (improvement minutes [95% Confidence Interval], 35.4 [24.2 to 46.6], and 20.6 [10.6 to 30.6] respectively), and support gave a higher remission rate (defined as ISI score below 8; 61.4%), than bibliotherapy alone (24.4%, p's < .001). Improvements were not seen in the control group (sleep onset latency 4.6 minutes shorter [-1.5 to 10.7], and remission rate 2.3%). Self-help groups maintained gains at three-month follow-up.ConclusionsParticipants receiving self-help for insomnia benefited markedly. Self-help, especially if therapist-supported, has considerable potential to be as effective as individual treatment at lower cost, also for individuals with co-morbid problems.Trial RegistrationClinicalTrials.gov: NCT01105052