The association between sleep duration and computerized neurocognitive testing outcomes at baseline and post injury in US service academy cadets and midshipmen: findings from the CARE Consortium

Purpose: The purpose of this study was to examine the relationship between sleep duration and computerized neurocognitive testing used to manage concussions among service academy cadets and midshipmen. Methods: A retrospective cohort study was conducted among participants enrolled at four US military service academies from 2014 to 2022. Participants underwent baseline concussion testing and reported demographic information during their first year at their respective academy. Participants who sustained a concussion underwent standardized post-injury evaluations throughout their recovery which included computerized neurocognitive tests (CNT). Self-reported sleep duration from the prior night was collected before each evaluation. Univariate and multivariable linear regression models were used to examine the relationship between sleep duration and CNT scores while controlling for significant covariates. Two-way ANOVA models were used to examine mean differences in CNT scores by hours slept and evaluation timepoint. Results: During the study period, 21,938 participants (26% female; 18.87 ± 1.34 y) enrolled and 1429 (36% female; 19.80 ± 1.50 y) completed a post-concussion CNT. Univariate linear regression models displayed significant associations between sleep and baseline scores of verbal memory (β = 0.024; 95% CI = 0.01, 0.04; p = 0.003), visual memory (β = 0.045; 95% CI = 0.03, 0.06; p < 0.001), visual motor speed (β = 0.052; 95% CI = 0.04, 0.07; p < 0.001), reaction time (β = -0.022; 95% CI = − 0.02, − 0.02; p < 0.001), simple reaction time (β = 0.094; 95% CI = 0.04, 0.15; p = 0.001), match to sample (β = 0.071; 95% CI = 0.01, 0.13; p = 0.013), and simple reaction time 2 (β = 0.110; 95% CI = 0.05, 0.17; p < 0.001). An association between sleep and specific CNT scores was observed at specific post-injury evaluations as well in univariate models. Multivariable linear regression models yielded similar results. Conclusion: Sleep increases were associated with improved CNT scores. Sleep duration the night before a baseline or post-concussion assessment should be considered when examining CNT results.