Predictors of Symptom Resolution Following Sport-Related Concussion

Journal of Athletic Training. -


Resch, J. E., H. R. Lee, C. N. Brown, T. A. Baumgartner, S. Olejnik, K. Walpert, S. N. Macciocchi and M. S. Ferrara.



Context: Resolution of sport-related concussion symptoms typically takes 7 to10 days. An athlete’s recovery may deviate significantly due to multiple factors such as prior history of concussion and current concussion severity. While there is variability in symptom resolution following concussion, a prediction equation may assist athletic trainers in estimating time until self-reported asymptomatic (SRA) following a concussion. Objective: Examine a multivariate regression model to predict SRA in a collegiate athlete sample. Setting: Athletic Training Laboratory Design: Cross-sectional study from the 2004 – 2009 sport seasons. Patients or Other Participants: Forty-four collegiate athletes diagnosed as concussed (32 males, 12 females): age 19.81 + 1.6 years. Interventions: Participants were evaluated 24 hours post-injury utilizing a self-reported symptom inventory, ImPACT, and the Neurocom Sensory Organization Test (SOT). Time until SRA was defined as number of days between athletes’ 24 hours post-concussion assessment and when athletes reported asymptomatic. Multiple regression analysis was performed utilizing the stepwise selection method. Significant relationships between predictors and the criterion were determined with = .05. Main Outcome Measures: Twenty-two self-reported symptoms ranked on severity and duration; composite, somatosensory, vestibular, visual, and visual conflict scores from the SOT; and the ImPACT composite scores visual and verbal memory, visual motor reaction time, reaction time, impulse control and symptoms were used as predictors, and time until SRA as the criterion. Results: Average SRA was 8.50 + 3.9 days. The linear model was time until SRA = 15.968 + (-.128 x ImPACT composite score visual memory) + (.927 x severity of sensitivity to noise) + (.737 x severity of blurred vision) + (-.915 x duration of irritability) + (-2.089 x nervousness) which was statistically significant, (R=.749,,R2 =.499,F =9.175 The strongest predictor to criterion relationships were between the ImPACT subscore visual memory (R = -.437, R2=.190, R2 = .191, P < .001), change severity of blurred vision (R = .420, R2=.176, R2 = .070, P < .001) and sensitivity to noise The weaker predictor to criterion relationships were duration of irritability and nervousness Conclusions: Overall, this prediction model accounted for approximately 50% of variance associated with time until SRA. Given its components; the current model supports the use of a battery of tests to assess sport-related concussion. The current model, although specific to the University of Georgia concussion paradigm may be used to estimate time until SRA in individual athletes, but cross-validation of the current model with a larger sample of concussed collegiate athletes is warranted.

Links to full article: