Hi, I'm Theo Hanna, a third-year medical student at Mount Sinai in New York, and I'd like to thank AANS and CNS for the opportunity to present my research on sex differences in the incidence, severity, and recovery of concussion in adolescent student-athletes. More females are playing football than ever before. It has been shown that in most sports, females are more susceptible to concussion and that football has the highest risk for concussion of any sport. This suggests that females playing football are at the highest risk for concussion. Our understanding of sex differences in concussion severity and recovery is less robust due in part to the lack of quantitative metrics, but developing a better understanding of sex differences in concussion recovery and severity is becoming increasingly important. The severity index was published by our research group this year and is based on impact data. In the study, we used it to evaluate sex differences in concussion severity and recovery in adolescent student-athletes. Our database includes over 11,000 baseline tests from student-athletes ages 12 to 22 between 2009 and 2019, over 5,000 of the subjects had post-injury testing that could be used to assess severity and recovery, we used regression analyses to assess sex differences in incidence and severity, and Cox hazard regressions to assess differences in recovery. Concussion was defined in the study as two out of five of the impact composite scores being significantly deviated from baseline based on previously published reliable change indices at the 80% confidence interval. This has been the standard practice in impact studies. Here recovery was defined as a return to zero or one composite score significantly deviated from baseline. The severity index calculation is shown here, but briefly it is the sum of the difference between post-injury and baseline score standardized by the reliable change index for each significantly deviated composite score. It was shown to correlate well with recovery time. There were some differences in demographics, with males more likely to play football be diagnosed with ADHD and have a history of two or more prior concussions. All five of the impact composite scores differed between males and females at baseline. Females had more symptoms but outperformed males on verbal memory processing speed and reaction time tasks. Males did better on visual memory tasks. We found no significant difference in the unadjusted incidence of concussion between males and females. Notably, however, females had greater deviations from their baseline and symptom score and processing speed following concussion than did males at their initial post-injury test. These differences in symptom score and processing speed were preserved as second follow-up testing after a concussion, despite females taking longer to return for their second follow-up test. We then performed multivariable analyses to control for demographic differences between the cohorts, including age, sport, ADHD, diagnosed learning disability, and prior concussion history. When controlling for these possible confounders, we found a significant effect of concussion on incidence, with females more likely to suffer concussion with an odds ratio of 1.62. Females also had higher severity index at first post-injury testing when controlling for demographic differences. Although the Kaplan-Meier curve and unadjusted Cox regression showed that females took longer to recover than males, when we controlled for initial concussion severity, this effect actually disappeared, as shown by the non-significant p-value in the multivariable Cox regression. In summary, females had significantly higher incidence of concussion, as has been previously described, and this effect was not just a result of increased symptom reporting, as some have previously speculated, because the results here also showed persistent deficits in the processing speed test and the objective portion of the neurocognitive assessment. Females had higher severity concussions, which were clinically significant, since they had protracted recovery compared to males in their unadjusted analyses. However, when we did control for initial concussion severity, there was no significant difference in recovery time, suggesting that severity and not sex is a major determinant of recovery time. This study is limited by its retrospective nature. Recovery length may be overestimated, since patients could have recovered any day prior to their second follow-up test. The recovery analyses in this study are not robust to symptom relapse during return-to-play protocols. Furthermore, it is not known how long it took participants who were still concussed at the end of our primary endpoint to recover. Additionally, impact is an imperfect measure of concussion and is susceptible to false patterns, positives and negatives.

Theo Hanna

medical student

sex differences

concussion

adolescent student-athletes