This study found no added benefit of prescribed EA or mHealth interventions over UC for acute concussion in youth. Although our findings did not support a positive effect of EA, multiple RCTs have shown that controlled, subsymptom threshold aerobic exercise initiated within 24-48 hours can facilitate recovery.^6,8 These studies emphasize that structured aerobic exercise, rather than general physical activity, is the key factor, with causality well established, though the underlying mechanisms remain unclear.³¹ Current guidelines recommend gradual return to activity as symptoms improve, a strategy supported by trials favoring subsymptom threshold exercise over passive rest. However, prior studies often involve small sample sizes (16-118 participants), clinic-based recruitment, and a focus on subacute or chronic recovery, limiting applicability to the acute phase.^32–34 Few RCTs have specifically evaluated exercise interventions within the first 72 hours postinjury in youth. In a 2019 RCT by Leddy et al,⁵ 103 adolescents seen in 2 community sports medicine clinics within a week of sport-related concussion were assigned to 20 minutes of daily subsymptom threshold aerobic exercise or a placebo-like stretching program. Exercise intensity was individualized using the Buffalo Concussion Treadmill Test, targeting 80% of the heart rate threshold for symptom exacerbation. The aerobic exercise group recovered significantly faster compared with the stretching group. A 2021 follow-up RCT replicated these findings across multiple sites with an intention-to-treat analysis, confirming that early (within 10 days) tailored aerobic exercise accelerated recovery and reduced risk of persistent symptoms.³⁵ Our study differed in several ways. We recruited from pediatric EDs within 72 hours of injury and provided standardized activity goals regardless of symptom status, rather than prescribing individualized, subsymptom threshold exercise. In contrast to the structured protocols used in prior trials, our general activity recommendations were unsupervised and not symptom-guided. A similar ED-based trial by Ledoux et al³⁶ randomized 456 youth to resume activity at 72 hours postinjury or follow UC. Although the primary analysis showed no group differences at 2 weeks, a per protocol analysis revealed that those who adhered to activity recommendations had fewer symptoms. Adherence, therefore, may be key. In our trial, although the activity group increased average daily step counts, only 38% met target step goals. This high nonadherence likely limited potential benefits and raises questions about the practicality of broadly recommending EA in acute care. Unlike structured, supervised exercise programs with tailored, subsymptom activity, both our study and Ledoux et al relied on general activity guidance without supervision or adjustments, which may explain the variability in outcomes. Additional differences, such as patient population (acute ED vs outpatient), timing of intervention (hours vs days postinjury), broader inclusion criteria, and differing outcome measures, may have contributed further to contrasting findings. Notably, our activity group transiently reported increased symptoms, which resolved by day 14, while the UC group showed better balance scores at follow-up. Together, these findings suggest that while physical activity is an essential component of concussion recovery, its benefits depend on timing and individualization. Prescribing a fixed activity level shortly after injury, without adjusting for symptom response, may be suboptimal. Current evidence supports tailored symptom-guided subsymptom threshold exercise after a brief period of relative rest.³⁷

Regarding the behavioral intervention, our commercially available mHealth app self-guided program showed no benefit, with only half the patients engaging with the app for a median of 5 days and meeting resilience goals. This limited acceptability suggests a need for more engaging or personalized digital tools. In addition, our mHealth programming lacked concussion-specific goals, symptom monitoring, recommendations, or direct health care team communication. While this approach did not show benefit, the app also did not cause harm, which contrasts with evidence suggesting increased screentime postpediatric mTBI may worsen symptoms.³⁸ Future concussion-specific programs integrating symptom-guided return to activity could prove beneficial. A planned subanalysis of a high-risk populations and a per protocol analysis may clarify further the benefits of prescribed activity or behavioral intervention.

Our study supports the current standard of care for acute concussion management. mTBI triggers a neurometabolic cascade leading to a brain energy crisis and functional impairments.³⁹ Rest theoretically restores brain energy stores. In our prior trial comparing prolonged rest (5 days) with UC (24-48 hours), prolonged rest worsened symptom outcomes without affecting neurocognitive outcomes.³ DiFazio et al⁴⁰ noted psychological and physical deconditioning effects contributing to symptom persistence in youth post-mTBI. Our study found 24-48 hours of rest followed by symptom-guided return to activity led to faster symptom resolution compared with early, symptom-independent activity. Combined with existing literature, these findings suggest personalized concussion guidance postinjury could prevent excessive rest or premature return to activity. Most patients can naturally increase activity as symptoms improve over 10-14 days postinjury. For those with variable symptoms or slow recovery, early referral to primary or specialty care may ensure smooth recovery and timely interventions. Future studies should focus on tailoring interventions to individual symptom thresholds, such as developing methods to determine personalized subsymptom threshold activity goals in acute care settings or utilizing remote patient monitoring to provide timely symptom-guided recommendations.

Our study was not without limitations. This trial encountered challenges common to studies conducted during the COVID-19 pandemic. Recruitment and retention difficulties prompted adaptations such as enhanced compensation and flexible follow-up options (eg, in-home and remote assessments), which were necessary for social distancing but limited our ability to capture secondary outcomes. Changes in care-seeking behavior during the pandemic may also have influenced our sample. Increased loss to follow-up may have introduced attrition bias, impacting the robustness of findings. Requiring smartphone ownership excluded 7% of eligible participants. Outcome assessors were not blinded, which may have introduced bias. Most participants were recruited from the ED or a specialty clinic within 72 hours of injury, whereas many adolescents with concussion present to primary care days to weeks later, limiting generalizability. Symptom resolution was self-reported and not confirmed with objective balance or oculomotor testing, potentially missing residual deficits. Adherence to early physical activity recommendations was low, and most injuries occurred in nonsport settings, in contrast to RCTs showing benefit in athletic populations.

Overall, our study found no evidence that early prescribed activity or mHealth interventions added benefit over standard care for youth with acute concussion. While prior work, including Ledoux et al’s per protocol analysis, suggests that individualized goal-directed exercise may support recovery, high nonadherence in our study limits conclusions about efficacy. These findings raise concerns about universal recommendations to resume physical activity within 72 hours postinjury, regardless of symptom burden. Instead, they support an individualized approach that emphasizes early, light physical activity below the symptom-exacerbation threshold to avoid both excessive rest and premature return to full activity. Despite 2022 Amsterdam Consensus guidance recommending on light activity within 24 hours, real-world adherence remains a barrier.² Future work should refine strategies to improve both effectiveness and uptake of tailored interventions.