Abstract Details

Approximately 50% of individuals who sustain a TBI develop chronic insomnia, with known deleterious consequences. The underlying physiologies and temporal trajectories associated with TBI-related insomnia are unknown.

To examine the relation between acute biomarkers of neuronal injury and trajectories of insomnia following traumatic brain injury (TBI)

Blood biomarkers (GFAP, UCHL1, hsCRP) and self-reported insomnia symptoms were collected from 1700 patients enrolled in the FITBIR-standardized multi-center TRACK-TBI study. Biomarkers were collected at post-injury days 1 and 14 (D1, D14). Insomnia Severity Index (ISI) was scored at post-injury D14, months 3, 6, and 12. Unconditional latent class mixed models (LCMMs) were constructed to model distinct ISI trajectory classes. Multinomial logistic regression models evaluated whether acute biomarker levels could differentiate the identified trajectory classes after adjusting for age, sex, race, psychiatric history and CT status.

LCMMs yielded a best-fit model of 5 latent classes of ISI symptom trajectories (C1=sustained severe; C2=sustained mild; C3=mild deteriorating; C4=severe resolving; C5=mild resolving). Adjusted regression models demonstrated that higher GFAP and lower hsCRP at D1 and/or D14 were associated with increased likelihood of membership in a lower-severity ISI class. Specifically, D1 hsCRP significantly differentiated C1 versus C2 (OR=1.2 [1.02-1.42] per log unit decrease). Both D1 and D14 GFAP and hsCRP differentiated C1 versus C5 (D1 GFAP OR=1.39 [1.15-1.67] per log unit increase & hsCRP OR=1.37 [1.15-1.63] per log unit decrease; D14 GFAP OR=1.59 [1.09-2.33] per log unit increase & hsCRP OR=1.39 [1.09-1.72] per log unit decrease). D14 hsCRP additionally differentiated C1 from C4 (OR=1.43 [1.02-2.0] per log unit decrease). UCHL1 levels were not significantly associated with trajectory likelihood.

Following TBI, acute levels of GFAP and hsCRP predict trajectories of insomnia symptoms, including persistent severe insomnia. Future studies are needed to examine the underlying mechanisms and clinical implications of these associations.