Effects of hours of sleep on ImPACT concussion testing: comparing baseline with postinjury scores

Eugene I Hrabarchuk,Roshini Kalagara,Bahie Ezzat,Frederika Rentzeperis,Tomasina M Leska,Alexander J Schupper,Benjamin Rodriguez,Muhammad Ali,Addison Quinones,Lily McCarthy,Mathew T Carr,Arielle B Lehman,Alex Gometz,Mark Lovell,Tanvir F Choudhri

DOI: https://doi.org/10.3171/2024.2.PEDS23437

2024-05-03

Abstract:Objective: The influence of sleep on baseline and postconcussion neurocognitive performance prior to Immediate Post-Concussion Assessment and Cognitive Testing (ImPACT) is poorly understood. Since ImPACT is widely used in youth sport to assess neurocognitive performance before and after head injury, it is important to delineate factors that affect testing performance. While some have reported correlations between fewer hours of sleep and lower scores on baseline tests, others have not observed any such associations. Therefore, the authors sought to compare the relationship between sleep and neurocognitive performance on ImPACT at both baseline and postinjury. Methods: The authors queried a database of 25,815 ImPACT tests taken from 2009 to 2019 by athletes aged 12-22 years. There were 11,564 baseline concussion tests and 7446 postinjury concussion ImPACT tests used in the analysis. Linear regression was used to model the effect of sleep on baseline and postconcussion ImPACT scores adjusting for sex, age, learning disability, attention-deficit/hyperactivity disorder, number of prior concussions, number of games missed, and strenuous exercise before testing. Results: Mean composite scores expectedly were all significantly lower in the post-head injury group compared with the baseline group. In the multivariable analysis, at baseline, hours of sleep significantly affected symptom scores (β = -1.050, 95% CI -1.187 to -0.9138; p < 0.0001). In the postinjury multivariable analysis, verbal memory (β = 0.4595, 95% CI 0.2080-0.7110; p = 0.0003), visual memory (β = 0.3111, 95% CI 0.04463-0.5777; p = 0.0221), impulse control (β = -0.2321, 95% CI -0.3581 to -0.1062; p = 0.0003), and symptom scores (β = -0.9168, 95% CI -1.259 to -0.5750; p < 0.0001) were all affected by hours of sleep. Conclusions: Hours of sleep did not alter neurocognitive metrics at baseline but did have an impact on post-head injury metrics. These findings suggest that individuals may be able to compensate for lack of sleep at baseline but not immediately after concussion. Concussions may reduce cognitive reserve or detract from the brain's resources, making sleep even more important for proper neurocognitive functioning postconcussion. Future work will analyze the effects of sleep on postconcussion test performance.