Influence of Postconcussion Sleep Duration on Concussion Recovery in Collegiate Athletes.

Objective: To determine whether decreased sleep duration postconcussion influences days to asymptomatic and assessment of performance throughout recovery.

Design: Prospective.

Setting: Institutional Clinical Research Laboratory.

Patients: Four hundred twenty-three collegiate athletes were diagnosed with concussion.

Interventions: Multidimensional concussion assessment battery was conducted at baseline, within 24 to 48 hours, daily [2-4 days postinjury (PI); symptoms only], once asymptomatic, and after return-to-play. The battery included the following: 22-item symptom checklist, Standardized Assessment of Concussion (SAC), Balance Error Scoring System (BESS), and computerized neurocognitive test [Immediate Post-Concussion Assessment and Cognitive Testing (ImPACT)].

Main Outcome Measures: We subtracted baseline sleep duration from 24 to 48 hours postconcussion sleep duration and categorized athletes into the following groups: shorter sleep (≤-1 hour), no change (>-1 hour, <+1 hour), and longer sleep (≥+1 hour). A 1-way analysis of variance (ANOVA) was conducted to compare days to asymptomatic and separate mixed-model ANOVAs to compare total symptom scores, SAC total scores, BESS total error scores, and ImPACT composite scores between sleep categories across time points (α = 0.05).

Results: Sleep groups did not differ in days to asymptomatic. The shorter sleep group had greater symptom severity than no sleep change and longer sleep groups at 24 to 48 hours (shorter: 39.1 ± 20.7; no change: 25.1 ± 18.4, P = 0.007; longer: 25.7 ± 21.8, P = 0.004), and at 2 to 4 days PI (shorter: 21.8 ± 21.8; no change: 10.5 ± 10.8, P = 0.013; longer: 11.9 ± 14.2, P = 0.007), but did not differ at other time points (ie, asymptomatic and return-to-play). Participants with shorter sleep exhibited slower ImPACT reaction times at 24 to 48 hours (shorter: 0.68 ± 0.14; no change: 0.61 ± 0.09, P = 0.016; and longer: 0.62 ± 0.12, P = 0.028) and asymptomatic time points (shorter: 0.62 ± 0.11; no change: 0.56 ± 0.05; P = 0.015).

Conclusion: Postinjury sleep declines may be associated with symptom severity and worsened reaction time during initial stages of recovery or may be the result of the concussion itself. Clinicians should be aware of alterations in sleep duration and manage appropriately to mitigate initial symptom burden postconcussion.