ADOLESCENT CONCUSSION: POST-INJURY ASSESSMENT RELATIONSHIPS
Rationale/Conceptual Basis/Background: Post-injury symptom assessment and neurocognitive testing are important components of the multi-dimensional approach to concussion recognition and management prevalent in current concussion guidelines. Previous studies have found mixed results related to individual factors that may affect post-injury assessment and sequelae of this type of head injury.
Methods: Study Design and Protocol: This exploratory investigation is a secondary data analysis of an established post-injury dataset. High School athletes across the United States completed a computer-based assessment that included a self-report demographic and concussion history questionnaire, the Post-Concussion Symptom Score (PCSS) self-report symptom scale, and specific neurocognitive tests that comprise the post-concussion assessment battery known as ImPACT. A random sample of 1515 de-identified adolescents aged 14-18 years was extracted from an overall dataset of 6983 with 69.1% male, average age and grade level of 15.9 years and 9.34, respectively. BMI mean =36.8, concussion mean =1.2, with those resulting in LOC = 0.4. Sports most commonly represented included football (41.6%) and soccer (15.4%). Data Analysis: Hierarchical multiple regression was conducted to determine the relevant contribution of each independent variable (age, gender, BMI, sport played, identified cognitive risk factors, and number of previous concussions with or without LOC) to each dependent variable using 3 established symptom and 5 established neurocognitive composites. Block entry was performed with Cognitive Risk Factors as 1st block, Demographics as 2nd block and Concussion Specifics as the final block, order of entry supported by theory and previous research findings.
Results: Results of the Symptom regression resulted in an overall model that was statistically significant for the composites of Physical/Neural (p=.025) and Emotion (p<.001). The symptom composite, Sleep, was not statistically significant. Results of the Neurocognitive regression resulted in an overall model that was statistically significant for the composites of Verbal memory (p=.007), Visual Memory (p=.001], Visual Motor (p =.001) and Impulse Control (p<. 001). The neurocognitive composite, Reaction Time, was not statistically significant. Variance accounted for in addition to results of individual variables will be presented in greater detail.
Implications: This study sought to further examine relationships to concussion as theoretical blocks. Despite results that indicated significant relationships, they were few and those that were significant accounted for low variance. Although important findings for recognition and management of concussion, the findings likely represent the diversity of the dataset that contained post-injury adolescents with an average post-injury time of 21.88, representing athletes at various stages of recovery.