A randomized controlled intervention of workplace-based group cognitive behavioral therapy for insomnia

PURPOSE: Sleep disturbance is common in the working population, often associated with work stress, health complaints and impaired work performance. This study evaluated a group intervention at work, based on cognitive behavioral therapy (CBT) for insomnia, and the moderating effects of burnout scores at baseline. METHODS: This is a randomized controlled intervention with a waiting list control group. Participants were employees working at least 75% of full time, reporting self-perceived regular sleep problems. Data were collected at baseline, post-intervention and at a 3-month follow-up through diaries, wrist-actigraphy and questionnaires including the Insomnia Severity Index (ISI) and the Shirom–Melamed Burnout Questionnaire (SMBQ). Fifty-one participants (63% women) completed data collections. RESULTS: A multilevel mixed model showed no significant differences between groups for sleep over time, while there was a significant effect on insomnia symptoms when excluding participants working shifts (N = 11) from the analysis (p = 0.044). Moreover, a moderating effect of baseline-levels of burnout scores was observed on insomnia symptoms (p = 0.009). A post-hoc analysis showed that individuals in the intervention group with low burnout scores at baseline (SMBQ < 3.75) displayed significantly reduced ISI scores at follow-up, compared to individuals with high burnout scores at baseline (p = 0.005). CONCLUSIONS: Group CBT for insomnia given at the workplace did not reduce sleep problems looking at the group as a whole, while it was indicated that the intervention reduced insomnia in employees with regular daytime work. The results also suggest that workplace-based group CBT may improve sleep in employees with primary insomnia if not concomitant with high burnout scores. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1007/s00420-018-1291-x) contains supplementary material, which is available to authorized users.