This study was approved by the Hamilton Integrated Research Ethics Board (REB #14‐376). Informed assent/consent was obtained from participants and parents.

Participants were recruited from the local Hospital Emergency Department, community referrals from their primary physician, and rehabilitation or sports medicine clinics. Eligibility criteria included the following: a physician-diagnosed concussion within the past 12 months, being 5‐18 years of age, active symptomatology, and English-speaking ability. Youth were deemed ineligible if they had a confirmed brain injury requiring resuscitation, admission to the pediatric critical care unit, or surgical intervention, and if they refused to wear the ActiGraph. This prospective longitudinal cohort study had 3 measurement time points: recruitment/first visit; symptom resolution; and final visit, which occurred 3 months post symptom resolution or 6 months post enrollment if symptoms did not resolve within the study time frame. This investigation consisted of various outcomes, including electroencephalogram, [19,20] magnetic resonance imaging (MRI) [21], cognition [22,23], and sleep [24], which were published previously. Data on self-reported adherence, and adherence to RTS protocols specifically, were published by DeMatteo et al [8].

The CanChild protocols [3,4] consist of 6 stages of RTA and 5 stages of RTS, made with reference to the Zurich guidelines [25]. Youth were advised that no high intensity physical activity or contact sports was permitted while they were symptomatic. They were also informed that “rest” does not equate to social isolation or sensory deprivation. Once recruited, youth received the ActiGraph and the 2015 CanChild protocols [3,4] immediately. Participants completed surveys every 48 hours using REDCap (Research Electronic Data Capture; version 14.5.10, 2024; Vanderbilt University)—a browser-based data management application [26]. The surveys included the Post-Concussion Symptom Scale (PCSS) [27], RTA/RTS stages, and an assessment of cognitive activity [3]. The PCSS [27] is common across concussion evaluations [7] and consists of a 22-symptom checklist scored on a 0‐6 Likert scale. This was adapted for younger children using a dichotomous yes/no scale [28]. The cognitive scale assessed cognitive activity on a scale of 1‐5 and was adapted from Brown et al [29]. The second in-person visit occurred at symptom resolution. The label of symptomatic or nonsymptomatic was based on the return of participants’ self-identified current reporting of symptoms to their preinjury symptom status.

To assess physical activity in the RTA protocols, youths were outfitted with an ActiGraph Gt3X waist-worn monitor accelerometer (ActiGraph LLc). The ActiGraph accelerometer provides a high-resolution measure of the duration, intensity, and frequency of movement and is validated for use in youth [12-17]. Participants were provided standardized instructions on how to wear the accelerometer and to record times of nonwear in a log diary. Accelerometry data were downloaded into 30-second epochs and visually inspected by trained personnel to ensure wear times matched those reported by participants. The data were cleaned to remove any nonwear periods or spurious data using ActiLife (version 6.13.4; ActiGraph LLc). The 30-second epoch was selected for analyses, as shorter epochs are more accurate to measure exercise intensity during intermittent physical activity [30]. Only valid days, defined as at least 6 hours and no more than 19 hours of wear time, were included in the analysis. Activity count data were then scored for analysis of adherence. To do this, daily time spent being sedentary or engaging in light physical activity (LPA), moderate activity, vigorous physial activity, and moderate-to-vigorous physical activity (MVPA) were calculated using the Evenson et al [14] cut points. Youth were considered adherent if there was 80% adherence to the physical activity requirements for the corresponding stage of the RTA protocol. Only participants who had complete actigraphy data were included in analyses (n=84). For stage 1, unlimited LPA was permitted, MVPA was limited to ≤2.5% of wear time and no consecutive bouts of ≥5 minutes at any intensity were recommended. In stage 2, baseline activity observed in stage 1 was permitted, as well as an extra 30 minutes of LPA, but no consecutive bouts ≥5 minutes of MVPA were recommended. In stage 3, baseline activity observed in stage 2 was permitted, as well as an extra 60 minutes of MVPA and upto two 15-minute bouts of MVPA. Only adherence to RTA stages 1‐3 were assessed with accelerometry because these stages had quantifiable activity amounts defined in the RTA protocol [4], and RTS did not have objective adherence data.

Subjective reported adherence to the RTA/RTS protocols was based on the following criterion: if participants received a label of “yes” to the questions asked by the research personnel: “Has the child been following the RTA/RTS guidelines correctly?” based on a self-reported progression through the RTA/RTS stages being associated with a decreasing self-report PCSS score. [8] Participants were categorized separately for RTA and RTS. Participants labeled as “Did not adhere” to both RTS and RTA, as well as participants labeled as “Adhered” to one but not both RTA and RTS protocols, were deemed “Did not adhere.”

Participants completed the short form of the Children’s Depression Inventory (CDI) [31]; the KIDSCREEN-52 [32]; Immediate Post-Concussion and Cognitive Test (ImPACT) [33]; and subsets of balance, bilateral coordination, running speed, agility, and strength from the Bruininks-Oseretsky Test of Motor Proficiency Second Edition (BOT-2) [34] at each in-person visit.

Demographic and injury data are presented as mean (SD), recruitment details are reported in percentages, and the PCSS score is reported as median, which better reflects the data due to a few outliers. Adherence for each participant was calculated for stages 1‐3 of RTA as described above. Adherence was determined a priori to be considered the primary predictor of outcomes, but as the adherence rate was very low, alternative analyses were performed (as explained in the Results section). The ActiGraph calculated adherence for RTA was then compared to the self-reported rating of adherence for RTA, and agreement was assessed using Cohen κ.

The rate of repeat head injury was calculated as a percentage of total injuries. A Mann-Whitney U Test was performed to assess time to RTA stage 3 and 6 and time to symptom resolution for those who adhered or did not adhere [15,35]. Only participants who reported continued symptoms up until 3 months post symptom resolution, self-reported adherence to RTA/RTS, and reported the final stage of RTA/RTS were analyzed. Participants who responded “no” to the question “On more than one occasion, have you had any symptoms of concussion in the last two weeks?” were given a PCSS score of 0. Significance was set at P=.05. Scores for the BOT-2, CDI, ImPACT, and KIDSCREEN-52 were reported as mean (SD) and median. All the data were tested for normality using the Shapiro-Wilk test. Time to symptom resolution, time to stages, and PCSS scores were not normally distributed. Data were analyzed using SAS (version 9.4; SAS Institute) and SPSS (version 23.0; IBM Corp), with significance set at P<.05.

Of the 139 participants who consented to the study, 107 (76.9%) participants completed follow-up assessments, 12 (8.6%) participants were lost to follow-up, and 20 (14.3%) participants withdrew from the study. Of the 20 participants who withdrew, 7 (35%) participants did so immediately after consent, 12 (60%) participants withdrew after the first in-person visit, and 1 (5%) participant withdrew before the final visit.

The cohort included 64 (46%) boys and 75 (54%) girls with a median age of 13.4 years. A total of 103 (74.1%) participants sustained their concussion via a sports-related injury, with most injuries obtained during recreational play (n=29, 28%) and ice hockey (n=26, 25%). This was the first concussion for 58.3% (n=81) of participants (Table 1). The median time from injury to the first visit was 7.8 days (mean 34.8 days, minimum 2.9 hours, and maximum 320.9 days). The mean time from injury to symptom resolution visit was 95.4 (SD 43.4) days for first in-person visits and 162.6 (SD 75.7) days for final visits.

Of the participants who remained in active enrollment (n=114), 16 (14%) participants did not achieve symptom resolution in the 6-month follow-up period (Table 1). Median time to symptom resolution was 16 days (Q1-Q3: 8-28; mean time 27, SD 33 days).

The rate of participants having another concussion during the follow-up period was 2% (n=3).

Based on the participant analysis, 13% (4/30) of participants adhered to stage 1; 11% (8/74) adhered to stage 2; and 34% (17/50) adhered to stage 3 (Table 2). Of note, only 1 participant from this cohort (N=139) adhered to all 3 stages.

ActiGraph data with sufficient wear time and the corresponding PCSS score and RTA stage were considered complete and then analyzed in 30-second epochs for 80% adherence to stages 1, 2, and 3. Participants were labeled as “Adhered” to each stage if they had at least 1 day in adherence to ActiGraph cut points for stage 1, 2, or 3. Participants were given a final label of “Did not adhere” if they did not meet the cut points corresponding to stage 1, 2, or 3.

Of the 105 participants with self-reported data, 59 (56.1%) participants adhered to the RTA protocol [11], 56 (53.3%) adhered to the RTS protocol, and 50.4% (n=53) were adherent to both protocols (Table 3).

Cohen κ was performed to determine if there was agreement between actigraphy and self-reported adherence to the RTA protocol. There was no statistically significant agreement between the two measures (κ=0.49, 95% CI 0.32‐0.66; P=.57; Table 4). Among the 84 participants with both ActiGraph and self-reported data, there was 48% (n=40) agreement between the two. A total of 36 (43%) participants self-reported adherence to the RTA protocol but failed to meet the ActiGraph adherence cut points.

Those with subjective reported adherence to the RTA protocol had a significantly shorter time in days (median=13) to symptom resolution than those who did not subjectively adhere (median=20; U=724.50; P=.03; Table 5).

The difference in time to symptom resolution was assessed, using a Mann-Whitney U test, for participants who self-reported adherence and nonadherence to RTA and RTS protocols (n=90). Time to symptom resolution was calculated as the time from initial injury to symptom resolution. Only participants who had a date of symptom resolution verified by research personnel were included in the analyses.

There was no statistically significant difference in time from injury to RTA stage 3 (P=.61) or stage 6 (P=.24) for participants who self-reported adherence or nonadherence (Tables 6 and 7).

The difference in the average number of symptoms as reported on the PCSS or postconcussion system inventory at stage 5 of RTS and stage 6 of RTA was assessed, using a Mann-Whitney U test, for participants who self-reported adherence and nonadherence to RTS/RTA protocols (n=86). There was no statistically significant difference in the PCSS score at stage 5 or stage 6 of RTS (P=.78) and RTA (P=.29), respectively, for participants who adhered or did not adhere to the protocols (Table 5).

The KIDSCREEN-52 physical and psychological well-being subsections scores improved from the first to final visit across most participants. The scores were considered “high,” demonstrating that participants felt they were physically fit and healthy and viewed life positively [29]. Participants’ CDI total T-score decreased for symptoms of depression from the first to the final visit (P=.33), where scores were in the average/low range (<60). Across all 3 visits, most participants scored in the “average” category. From the first to the final visit, the ImPACT subsection scores increased, suggesting an improvement in cognitive performance (P=.13).

There was no significant difference in the BOT-2, CDI, and KIDSCREEN-52 total or subsection scores between those who reported they subjectively adhered or those who did not adhere to the RTA and RTS protocols. There was a significant difference in the ImPACT Impulse Control Composite Score at the final visit for those who adhered to the RTA protocol (mean score 7.3, SD 5.1) versus those who did not adhere (mean 11.9, SD 11.7; P=.04).