Julie C. Wilson

Objective: To examine the association between brain neurometabolite concentrations and physical activity (PA) volume after adolescent concussion. Methods: We performed a longitudinal study of adolescents (13-18 years) with concussion. At an initial visit (visit 1; within 21 days of injury), participants rated their symptoms using the Post-Concussion Symptom Inventory (PCSI) and underwent magnetic resonance spectroscopy (MRS) to quantify neurometabolite concentrations in the posterior cingulate gyrus (PCG). Metabolites of interest were glutathione (GSH), glutamate-glutamine (Glx), total choline (tCho), and myo-inositol (mI). Following visit 1, participants were provided with wrist-worn activity monitors for the subsequent 14 days to track steps/day. They repeated MRS evaluation within 5 days of receiving medical clearance to return to sports (visit 2). We compared neurometabolite concentrations between visits, then used linear regression to investigate how PA was associated with neurometabolite concentrations at each visit independently. Results: Twenty-seven adolescents underwent MRS scans (15.8 ± 1.3 years; 56% female; assessed 11.7 ± 4.2 and 37.3 ± 16.1 days postinjury) and n = 16 completed PA monitoring. Neurometabolite concentrations did not differ between visits 1 and 2 ( P > .20) before accounting for PA. After controlling for symptom severity, glutamate-glutamine (β=407.5; 95% CI = -6.49, 821.47; P = .053) and total choline (β= 3953.79; CI = -38.47, 7946.05; P = .052) were marginally, but nonsignificantly, associated with PA volume. PA volume during the monitoring period was not significantly associated with unadjusted neurometabolite concentrations at visit 2 ( P > .40). Conclusions: Neurometabolite concentrations were marginally associated with subsequent PA volume. These findings indicate that MRS may provide useful information about how PA and brain neurochemistry are associated during concussion recovery.

Context : Neuromuscular training (NMT) is an effective injury prevention strategy but may have barriers to in-person delivery. Our objective was to examine injury incidence in the year following NMT intervention delivered virtually and to identify functional test performance associated with subsequent injuries. Design : Secondary analysis of randomized controlled trial. Methods : We recruited 40 physically active adults aged 18–40 years. Functional testing included simple (stimulus-response) and clinical (drop stick) reaction time (RT), single- and dual-task tandem gait, vestibular/oculomotor function, and cervical spine proprioception (head repositioning accuracy). Participants were randomized to virtual NMT or control groups. Virtual NMT was an 8-week program delivered via smartphone application, consisting of 3 self-guided sessions/wk. The main outcome was subsequent injuries sustained in the year following study completion. Results : Forty participants were randomized to intervention (n = 20) or control (n = 20). Thirty-five (87.5%) completed the intervention period. Twenty-four (69%) completed the 1-year follow-up injury survey (age = 26.0 [3.4] y, 71% female; n = 11 intervention and n = 13 control). Five participants reported a subsequent injury, and 19 did not. There was no proportional difference in subsequent injuries between the intervention (27.3%) and control groups (15.4%; P = .63). However, those who sustained subsequent injuries demonstrated significantly slower clinical RT (260.2 [18.7] vs 238.4 [17.3] ms, P = .02, Hedge g = 1.2) and greater head repositioning accuracy error (4.9 [1.8] vs 3.4 [0.8] degrees, P = .007, Hedge g = 1.5). Although not statistically significant, those who sustained injuries demonstrated slower tandem gait (single task: 17.0 [4.2] vs 14.7 [2.4] s, P = .12; dual task: 20.8 [6.2] vs 17.5 [4.1] s, P = .17) and more positive vestibular/oculomotor tests (2.0 [2.3] vs 0.7 [0.95], P = .06) compared to those who did not. Simple RT was similar between groups (467 [64.7] vs 457 [53.5] ms, P = 0.78). Conclusions : We did not observe a proportional difference in subsequent injuries between virtual NMT and control groups among uninjured adults. However, slower clinical RT and worse cervical spine proprioception may help identify individuals with an increased risk of injuries.

Context : Clinical assessments of the cervical spine and vestibular/oculomotor systems may inform return-to-play (RTP) clearance time after concussion and inform individualized concussion prognosis. Our objective was to investigate whether cervical spine and vestibular/oculomotor clinical assessments were associated with RTP clearance time after concussion. Design : Prospective cohort study. Methods : Adolescents ages 13–18 years within 3 weeks of concussion completed assessments of symptom severity using the Post-Concussion Symptom Inventory, cervical spine proprioception (head repositioning accuracy [HRA]), and symptom provocation with vestibular/oculomotor testing (visio-vestibular exam [VVE]; assesses smooth pursuits, saccades, vestibulo-ocular reflex, and motion sensitivity). HRA was performed with eyes closed and involved patients relocating their head to a neutral starting position after right/left rotation and cervical spine flexion/extension. The mean distance (distance between center position and each self-reported center point) across 12 trials was used in analysis as the measure of HRA error. For VVE, we used the number of tests that provoked symptoms for analysis. We then monitored participants until they received RTP clearance from their physician. We used multivariable linear regression with RTP clearance time (days from concussion to RTP clearance) as the outcome, HRA error and number of positive VVE subtests as predictors, adjusting for covariates. Results : We assessed 68 participants (age = 15.8 (1.4) y, 59% female, mean = 11.0 (3.8) d of postconcussion). After adjusting for age, sex, concussion history, initial symptom severity (Post-Concussion Symptom Inventory score), and days from concussion to assessment, neither HRA error (hazard ratio = 0.89; 95% confidence interval, 0.71–1.12; P = .34) or number of positive VVE subtests (hazard ratio = 1.06; 95% confidence interval, 0.91–1.24; P = .47) were associated with RTP clearance time. Conclusions : Cervical spine proprioception and vestibular/oculomotor symptom provocation evaluated within 3 weeks of injury were not associated with RTP clearance time, potentially reflecting the complexity and multifaceted nature of RTP decisions.

We examined associations between magnetic resonance imaging (MRI)–obtained cervical spine flexor and extensor muscle fat infiltrate and muscle volume with (1) return-to-play clearance and (2) postconcussion injury incidence. Adolescents (N = 30, age = 15.8 ± 1.3 years, 60% female) underwent cervical spine MRI initially postconcussion and were monitored until return to play. After return-to-play clearance, participants underwent follow-up MRI and completed monthly injury surveys for 12 months. Cox proportional hazards regression examined associations between muscle fat infiltrate and muscle volume (1) initially postconcussion with return-to-play clearance and (2) at return-to-play with subsequent injury incidence. Adjusting for covariates, return-to-play clearance was not associated with cervical spine muscle fat infiltrate (hazard ratio [HR] = 0.99, 95% confidence interval [CI]: 0.96-1.02) or muscle volume (HR = 0.98; 95% CI: 0.96-1.01). Adjusting for covariates, postconcussion injuries were not associated with muscle fat infiltrate (HR = 0.90; 95% CI: 0.78-1.05) or muscle volume (HR = 1.01; 95% CI: 0.97-1.04). Cervical spine muscle characteristics may not be affected after concussion, or our MRIs may have occurred too early to capture postconcussion dysfunction.

Context: Reaction time (RT) is a critical element of return to participation (RTP), and impairments have been linked to subsequent injury after a concussion. Current RT assessments have limitations in clinical feasibility and in the identification of subtle deficits after concussion symptom resolution.Objectives: To examine the utility of RT measurements (clinical drop stick, simple stimulus-response, single-task Stroop, and dual-task Stroop) to differentiate between adolescents with concussion and uninjured control individuals at initial assessment and RTP.Design: Prospective cohort study.Setting: A pediatric sports medicine center associated with a regional tertiary care hospital.Patients or Other Participants: Twenty-seven adolescents with a concussion (mean age = 14.8 ± 2.1 years; 52% female; tested 7.0 ± 3.3 days postconcussion) and 21 uninjured control individuals (mean age = 15.5 ± 1.6 years; 48% female).Main Outcome Measure(s): Participants completed the Post-Concussion Symptoms Inventory (PCSI) and a battery of RT tests: clinical drop stick, simple stimulus-response, single-task Stroop, and dual-task Stroop.Results: The concussion group demonstrated slower clinical drop stick (β = 58.8; 95% CI = 29.2, 88.3; P < .001) and dual-task Stroop (β = 464.2; 95% CI = 318.4, 610.0; P < .001) RT measures at the initial assessment than the uninjured control group. At 1-month follow up, the concussion group displayed slower clinical drop stick (238.9 ± 25.9 versus 188.1 ± 21.7 milliseconds; P < .001; d = 2.10), single-task Stroop (1527.8 ± 204.5 versus 1319.8 ± 133.5 milliseconds; P = .001; d = 1.20), and dual-task Stroop (1549.9 ± 264.7 versus 1341.5 ± 114.7 milliseconds; P = .002; d = 1.04) RT than the control group, respectively, while symptom severity was similar between groups (7.4 ± 11.2 versus 5.3 ± 6.5; P = .44; d = 0.24). Classification accuracy and area under the curve (AUC) values were highest for the clinical drop stick (85.1% accuracy, AUC = 0.86, P < .001) and dual-task Stroop (87.2% accuracy, AUC = 0.92, P < .002) RT variables at initial evaluation.Conclusions: Adolescents recovering from concussion may have initial RT deficits that persist despite symptom recovery. The clinical drop stick and dual-task Stroop RT measures demonstrated high clinical utility given high classification accuracy, sensitivity, and specificity to detect postconcussion RT deficits and may be considered for initial and RTP assessment.

We examined the association of self-reported sleep problems with clinical measures of postural stability, memory performance, symptom burden, and symptom duration following youth concussion. Patients 6-18 years of age presenting ≤21 days postconcussion underwent a clinical evaluation including modified Balance Error Scoring System, single- and dual-task tandem gait, immediate and delayed recall, and symptom severity. We calculated time from injury until symptom resolution and determined the proportion of patients who developed persistent postconcussion symptoms, defined as a symptom duration >28 days postconcussion. We grouped patients based on whether they reported sleep problems at their postconcussion clinical evaluation and compared symptom-based and functional outcomes between groups. Of the 207 patients included, n = 97 (14.3 ± 2.9 years; 49% female; initial visit 10.2 ± 5.8 days postconcussion) reported sleep problems postconcussion and n = 110 (14.3 ± 2.4 years; 46% female; initial visit 9.3 ± 5.4 days postinjury) did not. Those reporting sleep problems postconcussion had significantly more modified Balance Error Scoring System errors than those without (8.4 ± 5.5 vs 6.7 ± 4.7; P = .01), but similar tandem gait and memory performance. A significantly greater proportion of those who reported sleep problems postconcussion experienced persistent postconcussion symptoms than those who did not (53% vs 31%; P = .004). After adjusting for time from concussion to clinical visit and preconcussion sleep problems, postconcussion sleep problems were associated with a 2 times greater odds of developing persistent postconcussion symptoms (adjusted odds ratio = 2.02, 95% CI = 1.01, 4.06; P = .049). Identifying sleep problems early following concussion may allow clinicians to implement targeted treatment recommendations to improve sleep and provide an optimal recovery environment.

Context: The relationship between physical activity (PA) and fear of pain with movement (ie, kinesiophobia) during concussion recovery is unknown. Kinesiophobia may limit PA, while PA after concussion may reduce kinesiophobia. Our purpose was to examine the correlation between PA and self-reported kinesiophobia during concussion recovery for adolescents with and without persistent symptoms. Design: Prospective cohort study of children ages 10–18 years within 14 days of concussion. Methods: Participants rated kinesiophobia using the Tampa Scale of Kinesiophobia (TSK) at initial (≤14 d postconcussion) and return to play (RTP) assessments, and wore activity monitors between assessments. Our primary outcome was TSK score change from initial to RTP assessments. We grouped participants based on whether they experienced persistent symptoms (symptoms ≥28 days) or not (symptoms r for normally distributed and Spearman rho for nonnormally distributed variables) between PA variables and TSK change scores. Results: Among the 41 participants enrolled, 44% developed persistent symptoms (n = 18; age = 14.5 [2.0] y; 50% female; symptom duration = 57.3 [6.2] d; RTP = 66.8 [6.4] d) and 56% did not (n = 23; age = 14.9 [1.8] y; 48% female; symptom duration = 15.2 [1.5] d; RTP = 21.7 [1.9] d). For the persistent symptoms group, greater TSK change scores (mean = −2.5 [5.7] point change) were significantly and moderately correlated with higher daily step count (r = −.60, P = .008) and exercise frequency (r = −.63, P = .005), but were not correlated with exercise duration (ρ = −.12, P = .65). Among the no persistent symptoms group, TSK change scores (mean = −6.0 [5.0] point change) were not correlated with step count (r = −.18, P = .41) or exercise duration (ρ = .10, P = .67), and the correlation with frequency was low and not significant (r = −.34, P = .12). Conclusions: Regular PA during concussion recovery, regardless of duration or intensity, may help reduce kinesiophobia for those experiencing persistent symptoms.

Objective To examine the association between acute post-concussion dizziness, initial symptom severity, and postural stability with time to symptom resolution among adolescents. Background Identifying early post-concussion symptoms and functional deficits that predict symptom resolution can guide treatment strategies. Dizziness is among the most common concussion symptoms, and existing literature investigating the association between dizziness and recovery time is mixed. Design/Methods Participants underwent initial evaluation = 14 days post-concussion, and self-reported symptom severity using the Post-Concussion Symptom Inventory (PCSI). We used PSCI dizziness ratings to group participants: a difference between current and pre-injury dizziness ≥3 = dizzy; difference p = 0.009). After adjusting for potential confounding variables (initial symptom severity, mBESS tandem stance errors, TG time, and loss of consciousness) multivariable model results indicated initial symptom severity was the only variable associated with symptom resolution time (HR = 0.98; 95% CI: 0.96, 0.997; p = 0.025). Conclusions Total symptom severity, but not dizziness or postural stability, was significantly associated with symptom resolution time among adolescents following concussion. Individuals with moderate to severe post-concussion dizziness had higher average symptom scores indicating self-reported dizziness should be interpreted in the context of total concussion symptom burden.