Association Between Symptom Cluster Endorsement at Initiation of a Graduated Return-to-Activity Protocol and Time to Return to Unrestricted Activity After Concussion in United States Service Academy Cadets

Background: The endorsement of symptoms upon initiation of a graduated return-to-activity (GRTA) protocol has been associated with prolonged protocols. It is unclear whether there are specific symptom clusters affecting protocol durations. Purpose: To describe the endorsement of specific concussion symptom clusters at GRTA protocol initiation and examine the association between symptom cluster endorsement and GRTA protocol duration. Study Design: Cohort study; Level of evidence, 2. Methods: This study was conducted among cadets enrolled at 3 US service academies. Participants completed an evaluation upon GRTA protocol initiation. Participants endorsing symptoms were binarized based on 6 symptom clusters (cognitive, emotional, insomnia, physical, sensitivity, and ungrouped). The primary outcome of interest was GRTA protocol duration based on symptom cluster endorsement severity. Prevalence rates were calculated to describe symptom cluster endorsement. Kaplan-Meier survival estimates and univariate and multivariable Cox proportional hazards regression models were calculated for all 6 symptom clusters to estimate GRTA protocol duration while controlling for significant covariates. Results: Data from 961 concussed participants were analyzed. Of these, 636 participants were asymptomatic upon GRTA protocol initiation. Among the 325 symptomatic participants, the physical symptom cluster (80%) was most endorsed, followed by the cognitive (29%), insomnia (23%), ungrouped (19%), sensitivity (15%), and emotional (9%) clusters. Univariate results revealed a significant association between endorsing cognitive (hazard ratio [HR], 0.79; p = .001), physical (HR, 0.84; p < .001), insomnia (HR, 0.83; p = .013), sensitivity (HR, 0.70; p < .001), and ungrouped (HR, 0.75; p = .005) symptom clusters and GRTA protocol duration. Endorsing physical (HR, 0.84; p < .001) and sensitivity (HR, 0.81; p = .036) clusters maintained a significant association with GRTA protocol duration in the multivariable models. Conclusion: Participants endorsing physical or sensitivity symptom clusters displayed GRTA protocols prolonged by 16% to 19% compared with participants not endorsing that respective cluster after controlling for significant covariates.