Abstract
Objective: Prior psychometric research has identified symptom subscales for the Post-Concussion Symptom Scale (PCSS) based on confirmatory factor analysis (CFA), including cognitive, physical, sleep-arousal, and affective symptom factors. Study objectives included: (1) replicate the 4-factor PCSS model in a diverse sample of athletes with concussion, (2) test the model for invariance across race, gender, and competitive level, and (3) compare symptom subscale and total symptom scores across concussed groups with established invariance.
Setting: Three regional concussion care centers.
Participants: A total of 400 athletes who completed the PCSS within 21 days of concussion (64% boys/men, 35% Black, and 69.5% collegiate athletes).
Design: Cross-sectional.
Main Measures: A CFA tested the 4-factor model and measurement invariance testing was performed across racial, competitive level, and gender groups. Symptom subscales and total symptom severity scores were compared based on demographic groupings with established invariance.
Results: The 4-factor model fit well and strong invariance was established across all demographic categories, indicating symptom subscales could be meaningfully compared across groups. Black and White athletes differed on total symptoms (U = 15 714.5, P = .021, r = 0.12), sleep-arousal symptoms (U = 15 953.5, P = .026, r = 0.11), and physical symptoms (U = 16 140, P = .051, r = 0.10), with Black athletes reporting slightly more symptoms. Collegiate athletes reported greater total symptom severity (U = 10 748.5, P < .001, r = 0.30), with greater symptom reporting on the cognitive (U = 12 985, P < .001, r = 0.21), sleep-arousal (U = 12 594, P < .001, r = 0.22), physical (U = 10 959, P < .001, r = 0.29), and emotional (U = 14 727.5, P = .005, r = 0.14) symptom subscales. There were no significant differences by gender in the total symptom score or subscale scores. After controlling for time since injury, no racial differences persisted, but a significant difference by competitive level in physical symptom reporting (F = 7.39, P = .00, [eta]2 = 0.02) and total symptom reporting (F = 9.16, P = .003, [eta]2 = 0.02) remained.
Conclusion: These results provide external validation for the PCSS 4-factor model and demonstrate that symptom subscale measurements are comparable across race, genders, and competitive levels. These findings support the continued use of the PCSS and 4-factor model for assessing a diverse population of concussed athletes.