Background: Growing evidence indicates early exercise may improve symptoms and reduce clinical recovery time after concussion, but research examining collegiate student-athletes is scarce.

Objective: The aim of this study was to compare symptom recovery time, clinical recovery time, and persisting post-concussion symptom (i.e., symptoms = 28 days) prevalence by the timing of light exercise initiation before the graded return to play (RTP) protocol among concussed participants.

Methods: Collegiate student-athletes (n = 1228; age 18.4 ± 0.9 years; 56.5% male, 76.3% division I; 33.7% = 1 prior concussion) across 30 institutions enrolled in the CARE Consortium completed post-concussion assessments and were monitored over time. Symptom recovery (days from injury to symptom resolution) and clinical recovery (days from injury to return to play protocol completion) was determined by the student-athletes' clinicians. Student-athletes were categorized by timing of light exercise initiation. Early (< 2 days post-concussion; n = 161), typical (3-7 days post-concussion; n = 281), and late exercise (= 8 days post-concussion; n = 169) groups were compared with the no-exercise group (n = 617; i.e., did not exercise prior to beginning the RTP protocol) for all analyses. Multivariable Cox regression models with hazard ratios (HR) and survival curves and a multivariable binomial regression model with prevalence ratios (PR) compared recovery outcomes between exercise groups while accounting for covariates.

Results: Compared to the no-exercise group, the early exercise group was 92% more probable to experience symptom recovery (HR 1.92; 95% CI 1.57-2.36), 88% more probable to reach clinical recovery (HR 1.88; 95% CI 1.55-2.28) and took a median of 2.4 and 3.2 days less to recover, respectively. The late exercise group relative to the no-exercise group was 57% less probable to reach symptom recovery (HR 0.43; 95% CI 0.35-0.53), 46% less probable to achieve clinical recovery (HR 0.54; 95% CI 0.45-0.66) and took 5.3 days and 5.7 days more to recover, respectively. The typical exercise group did not differ in hazard for symptom or clinical recovery (p = 0.329) compared with the no-exercise group. The prevalence of persisting post-concussion symptoms in the combined sample was 6.6%. Early exercise had 4% lower prevalence (PR 0.96, 95% CI 0.94-0.99) and typical exercise had 3% lower prevalence (PR 0.97, 95% CI 0.94-0.99) of persisting post-concussion symptoms, while the late exercise group had an elevated prevalence (PR 1.11, 95% CI 1.04-1.18) compared with the no-exercise group.

Conclusion: Exercise < 2 days post-concussion was associated with more probable and faster symptom and clinical recovery, and lower persisting post-concussion symptom prevalence. When considering our findings and existing literature, qualified clinicians may implement early exercise into their clinical practice to provide therapeutic treatment and improve student-athlete recovery.