Insomnia and self-perceived disability in workers with delayed recovery after mild traumatic brain injury/concussion

Introduction: Insomnia is a common complaint among persons with mild traumatic brain injury (mTBI). However, its impact on recovery after mTBI/concussion has not been characterized. Clarifying the association between insomnia and self-perceived disability may serve the vital role in understanding rehabilitation outcomes, including return to work after the injury. Materials and methods: Adults with mTBI/concussion sustained at the workplace were recruited from the largest rehabilitation hospital in Ontario between May 2012 and May 2014. Demographic, occupational, and health status data were collected from questionnaires, insurer records, and clinical assessment at the time of recruitment. The Insomnia Severity Index (ISI) measured the primary independent variable and the Sheehan Disability Scale measured perceived disability outcome. This variable was highly skewed and therefore classified as "mild/moderate" and "marked/extreme" disability. Bivariate associations were assessed by two-sided t-tests and chi-square tests. A logistic regression model was fitted using variables selected from the literature. Model fit was assessed using the Hosmer-Lemeshow goodness of fit statistic, AIC criterion, and the ratio of the deviance value/degrees of freedom. Multinomial logistic models with several categorizations of the outcome were compared with the binary logistic model. Results: Of 94 workers (61.2% male) with mTBI/concussion, with a mean age of 45.2 +/- 9.9 years and a median time since injury of 197 days, the majority (56%)were onwork disability. Compared with those reporting mild/moderate disability, individuals reporting marked/extreme disability had more severe insomnia (p = 0.002), depression (p <0.0001) and greater pain (p <0.0001). In the fully adjusted logistic model, a one-point ISI increase was associated with a 1.16 (95% CI 1.03-1.30) greater odds of reporting higher disability. No other covariates remained significant. Multinomial models using disability categories based on tertiles and established cutpoints exhibited loss of precision, though the main point estimates remained similar. Conclusion: While insomnia was significantly associated with greater disability, depression and pain were not, suggesting that they may be partially explained by poor sleep. Increasing awareness of the impact of insomnia, assessment and treatment at the early stages after the injury is warranted.