Rest and Return to Activity After Sport-Related Concussion: A Systematic Review of the Literature

This systematic review was completed in accordance with the guidelines from the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA).¹⁶ 

An electronic search was conducted in 6 databases: the Cochrane Central Register of Controlled Trials, CINAHL, SPORTDiscus, Educational Resources Information Center, Ovid MEDLINE, and PubMed. The search terms were brain concussion AND academic accommodations, brain concussion AND cognitive rest, brain concussion AND rest, concuss* AND sports AND academic, concuss* AND sports AND rest, concuss* AND sports AND return to learn, concuss* AND sports AND return to play, mild traumatic brain injury AND rest, mild traumatic brain injury AND cognitive rest, mild traumatic brain injury AND return to activity, mild traumatic brain injury AND return to play, recurrent concussion AND sport (Table 2). We also performed hand searches for relevant articles from the reference lists of relevant papers. All searches were conducted from the date of inception of each database through October 7, 2015.

After removing duplicates, a 2-step process was used to identify appropriate articles for the review (Figure). Initial screenings of the titles and abstracts were completed independently by 3 of the authors (T.C.V.M., J.H.L., C.E.W.B.). A meeting was held to discuss any articles for which consensus on inclusion was not obtained. This was followed by a full-text review by the same 3 authors and an additional meeting to discuss the remaining articles for which consensus was not obtained. Articles were excluded if they did not meet the inclusion criteria.

Articles were included if they were published in English, were original research, and evaluated the use of, compliance with, or effectiveness of physical or cognitive rest after concussion or provided empirical evidence supporting the graded return-to-activity progression. Excluded articles consisted of narrative (clinical) reviews, editorials, critically appraised topics, commentaries, abstracts, animal research, studies of non–sport-related concussion, or original research that did not address the primary clinical questions of interest.

Articles were categorized according to the clinical question of interest. The study design; patient or participant sample; instrumentation or interventions used; outcome measures; main results; and conclusions were extracted, as appropriate, from each article and entered on a standard data-collection form. Because many of the studies included primary and secondary outcomes, we limited extraction of the outcomes and presentation of the results to those outcomes that fit the clinical questions of interest. Studies were assigned a level of evidence as outlined by the 2011 Oxford Centre for Evidence-Based Medicine scale.¹⁷ 

We critically analyzed the included studies to evaluate the patients, methods, and results. We summarized the conclusions using a qualitative synthesis of the findings. Articles were categorized into groups based on their ability to answer 1 of the primary clinical questions of interest: use of rest, rest effectiveness, compliance with recommendations, or outcomes after graded return-to-activity progression. If a study addressed more than 1 clinical question, it could be included in more than 1 area. Summary tables were developed to present the results. Because of the observational nature of many included studies and the heterogeneity of the patients, participants, interventions, or outcomes, we were unable to pool the data; therefore, a meta-analysis was not possible.

The literature search resulted in 2851 potential articles (Figure; Table 2). After we removed 1612 duplicates found across databases or with different search terms, we screened the remaining 1239 articles to meet the inclusion and exclusion criteria. After reading the title and abstract, we removed 1096 articles (194 narrative reviews, 34 commentaries, 8 editorials, 5 abstracts, 2 critically appraised topics, 1 unpublished dissertation, 68 not in English, 19 consensus statements, and 765 not relevant to the clinical questions), leaving 143 articles for full-text evaluation. Application of the exclusion criteria during the full-text review resulted in exclusion of 84 articles. Another 19 articles were excluded during the data-extraction process, leaving 40 articles from which data were extracted that were synthesized in the results. These consisted of 9 studies of use of rest, 10 studies of rest effectiveness, 17 studies evaluating compliance with guidelines, and 4 studies of return-to-activity outcomes.

Nine studies evaluated the use of rest and associated academic adjustments by health care providers (Table 3). The study designs were prospective cohort (n = 1), retrospective cohort (n = 2), retrospective chart review (n = 2), and cross-sectional survey (n = 5); 1 study used 2 research designs (retrospective chart review and cross-sectional design).

Three studies specifically addressed the use of cognitive-rest recommendations, and their findings were similar. Although 62% of physicians indicated understanding that cognitive rest should be part of the concussion-management plan, Arbogast et al¹⁸  reported that only 11% of adolescent medical records included written recommendations for cognitive rest. Similarly, Upchurch et al¹⁹  conducted a retrospective chart review and reported that cognitive-rest recommendations were not made to any patient before 2008, and recommendations for patients to rest increased only to 12% by 2012. Zemek et al²⁰  also noted that cognitive-rest recommendations provided by physicians were limited.

Two studies addressed the effect of a patient's medical history and postconcussion symptoms on cognitive rest and return-to-learn progressions. After a retrospective chart review, Carson et al²¹  concluded that adolescents with a history of concussion required more days of rest than those without a history. Additionally, 44.7% of patients had worsening symptoms when they prematurely progressed through a return-to-learn protocol. Similarly, Grubenhoff et al²²  observed that adolescents with persistent postconcussive symptoms (ie, >1 month) missed 50% more school days than adolescents without such symptoms. However, the percentage of patients who received academic adjustments did not differ between those with and those without postconcussive symptoms, suggesting underuse of academic adjustments for patients who would likely benefit.

Four studies addressed health care professionals' involvement in return-to-learn protocols. Wilkins et al²³  conducted a retrospective cohort chart review and found that instructions for a “return-to-think” progression increased from 24% to 98% after the implementation of standardized concussion guidelines by health care professionals in a sports concussion clinic. Three of the 4 studies used a cross-sectional survey to assess the role of school nurses or athletic trainers in implementing academic accommodations in the secondary school setting. Specifically, Weber et al²⁴  showed that 59% of student-athletes with concussions who were under the care of a school nurse received academic accommodations, whereas Olympia et al²⁵  demonstrated that 58% of school nurses were responsible for guiding the graduated academic-reentry process for student-athletes with concussions. Williams et al,²⁶  on the other hand, reported that 41% of student-athletes under the care of an athletic trainer employed in the secondary school setting received academic accommodations after a sport-related concussion.

Ten studies evaluated the effectiveness of rest in recovery (Table 4). The study designs and outcome measures had significant heterogeneity; some were focused on both cognitive and physical rest, and others only evaluated cognitive rest. Of the included studies, 4 were retrospective cohort, 3 were prospective cohort, and 3 were randomized controlled trials (RCTs). The outcome variables differed and included total symptom score, symptom duration, neurocognitive function, balance assessment, and clinical recovery, defined as a return to baseline on adjunct (cognitive and balance) assessments.

The findings varied: 2 studies^27,28  indicated that rest improved outcomes, 3 studies²⁹⁻³¹  identified too much activity as detrimental to recovery and resulting in worse outcomes, 4 studies³²⁻³⁵  found no association between rest and outcomes, and 1 study³⁶  showed that strict rest can lead to a longer recovery.

The authors^27,28  reporting that rest was beneficial to recovery used a similar clinical sample of patients presenting to a concussion specialty clinic. In a retrospective chart review of 13 patients with persistent concussion symptoms, patients were evaluated in the clinic 24.8 ± 30.7 days after their concussion and prescribed 1 week of cognitive and physical rest.²⁸  After the prescribed rest, patients had significant improvements in cognitive function and a reduction in the total symptom severity score. An earlier retrospective cohort study from the same clinic investigated how 1 week of prescribed cognitive and physical rest affected patients presenting with different durations of postconcussion symptoms.²⁷  After the week of rest, a significant improvement in cognition and reduction in symptoms were noted among all patients; no effect was noted for the length of time patients were symptomatic, which suggests that rest can be an effective treatment, regardless of whether it is prescribed acutely after concussion or is delayed.²⁷ 

In addition to these 2 studies, 3 other studies²⁹⁻³¹  showed that moderate levels of rest resulted in better outcomes compared with higher levels of activity, which may hinder recovery. In one of the first studies to evaluate rest and activity after sport-related concussion, Majerske et al²⁹  retrospectively analyzed how activity level influenced cognitive and symptom outcomes. The authors developed an activity intensity scale (AIS) that ranged from 0 to 4, with 0 indicating no school or exercise activity and 4 indicating school activity and participation in competitions. Using the AIS rating of 2 (school activity and light activity at home) as the reference, their analysis determined that AIS affected scores of visual memory and reaction time, with patients in the highest activity level (AIS = 4) having the worst visual memory and slowest reaction times.²⁹  However, no differences were noted between the lower AIS levels and the reference level, suggesting that moderate amounts of cognitive and physical exertion may be an appropriate management strategy.

Similarly, a prospective study that examined how cognitive activity levels affected duration of concussion symptoms found that patients engaged in the highest levels of activity took longer to resolve symptoms.³⁰  The authors developed a cognitive activity scale for patients to complete at each follow-up visit. The scale ranged from 0 to 4, with 0 indicating complete cognitive rest (no reading, homework, texting, video games) and 4 indicating full cognitive activity (no restrictions). Patients in the highest quartile of cognitive-activity days took significantly longer to recover than patients in the first to third quartiles, suggesting that too much activity may hinder recovery.³⁰  However, the recovery times of patients within the lower quartiles did not differ, indicating that refraining from all cognitive activity may not be necessary and that some activity may be beneficial to the recovery process.

A negative association between activity and recovery was also noted in a pilot RCT of collegiate athletes.³¹  Patients were randomized either to standard care (no exertion beyond normal school activities) or to an exertion group, who rode a stationary bicycle for 20 minutes at mild to moderate intensity and wore an ActiGraph device to track activity. Although the median days to recovery did not differ by group, the average amount of daily vigorous exertion was related to recovery: more vigorous activity resulted in a longer recovery time. Similar to other results, early mild to moderate exercise did not delay recovery, suggesting that it may be useful in managing patients with concussion.³¹ 

Four studies³²⁻³⁵  found no association between rest and outcomes. De Kruijk et al³⁵  evaluated the effect of 6 days of bed rest compared with no rest on patients presenting to an emergency department. Patients randomized to the bed-rest group reported less dizziness during the first 4 days after injury and a lower score for “feeling faint” at the 2-week follow-up, but other symptom scores and perceptions of quality of life did not differ between the groups at either time point, indicating that complete bed rest did not improve outcomes.³⁵  Similarly, in a retrospective cohort study of patients presenting to a sport concussion clinic, cognitive rest was prescribed to just over 46% of the patients; the recommendation for cognitive rest resulted in a longer mean symptom duration compared with patients to whom cognitive rest was not recommended.³²  However, once adjustments were made for other covariates, no association was noted between the rest recommendation and duration of symptoms.³²  These findings are similar to those reported by Moor et al,³⁴  who investigated adherence to rest and recovery time in adolescent athletes. Although patients generally followed the recommendations regarding physical and cognitive rest, adherence to rest was not a predictor of the length of time to recovery.³⁷  Lastly, authors³³  of a prospective study of collegiate athletes before and after a concussion policy change investigated the acute effects of 2 days (day of injury and 1 day after) of prescribed complete physical and cognitive rest on concussion-assessment measures, including cognition, balance, and symptoms. The group prescribed rest was symptomatic longer than the no-rest group, but no differences among groups were found on any of the other outcome measures.³³ 

In contrast, too much rest was detrimental to recovery in 1 RCT.³⁶  This study of patients with concussion presenting to a pediatric emergency department randomized patients into a usual-care group and a strict-rest group. Patients in the usual-care group were given verbal recommendations for activity from the treating physician. Patients in the strict-rest group were given discharge instructions for 5 days of strict rest at home that included no school, work, or physical activity. After the intervention, the strict-rest group reported a higher symptom severity score and had a slower resolution of symptoms compared with the usual-care group, but there were no meaningful differences on cognitive and balance testing, which suggested that restricting activity too much may result in unfavorable outcomes after concussion.³⁶ 

Researchers in 17 studies evaluated provider knowledge, attitudes, and beliefs regarding return-to-activity guidelines as well as compliance with return-to-activity recommendations made in position or consensus statements (Table 5). All of the studies were descriptive in nature and varied regarding the providers surveyed; therefore, some may have limited generalizability.

Five studies³⁸⁻⁴²  were cross-sectional surveys of athletic trainers; the majority of the participants were members of the National Athletic Trainers' Association. These studies highlighted deficiencies in the use of multifaceted assessments for return-to-play decisions with varying levels of compliance. No study demonstrated full compliance with all 3 recommended areas of concussion assessment (graded symptom checklist, neuropsychological testing, and balance assessment) for return to play.

Five studies⁴³⁻⁴⁷  were cross-sectional surveys of various types of physicians, including team physicians, pediatricians, family physicians, and members of the Child Neurology Society, and determined that knowledge of and compliance with concussion guidelines varied. The clinical examination was the most frequently cited method for making return-to-activity decisions; however, respondents differed in their preferred methods for evaluating athletes. These results emphasize the need for additional provider education related to return-to-activity decisions.

Two groups^48,49  evaluated a mixed sample of coaches, clinicians or physicians, and athletic trainers using a cross-sectional survey design. Baugh et al⁴⁸  found that although most schools had concussion-management plans, compliance with specific components was lacking. Kroshus et al⁴⁹  assessed clinicians' perceived pressure to return patients to play and observed that female providers and those supervised directly by the athletic department perceived higher levels of pressure from coaches.

In 5 studies,⁵⁰⁻⁵⁴  investigators used epidemiologic data, chart reviews, or cross-sectional surveys to identify the recommendations given to those with concussion and adherence to guidelines. Emergency department or trauma center chart reviews^51,54  indicated that children were discharged without adequate concussion-specific activity restrictions (eg, restriction from athletic participation, avoidance of activities that require attention to visual or auditory stimuli). When patients were surveyed, a lack of appropriate return-to-play advice was evident.^52,53  Using high school injury registry data, Meehan et al⁵⁰  found no association between the timing of the athlete returning to play and the type of provider who made the return-to-play decision. These authors did not look at details of the type of assessments used.

No studies specifically addressed the effectiveness of graded return-to-activity progressions in improving patient outcomes; however, 4 groups⁵⁵⁻⁵⁸ evaluated aspects of the Zurich guidelines (Table 6). Of those, 3 studies examined return to sport after a standardized protocol based on the Zurich guidelines.⁵⁵⁻⁵⁷  Echlin et al⁵⁵  investigated recovery duration in junior ice hockey athletes using a clinical evaluation guided by the Zurich return-to-play protocol. Postinjury management was guided by the Zurich protocol, resulting in an average time of 12.8 ± 7.0 days before clinical recovery and return to hockey.⁵⁵  Chermann et al⁵⁷  studied the use of a standard protocol based on the Zurich guidelines in rugby athletes with concussions and reported a median of 6 days until symptoms resolved, with female athletes taking longer than male athletes. The mean number of days until patients returned to sport was 21 (range, 7–45 days), even after the mandatory International Rugby Board rule of removal from play for 3 weeks was phased out in 2011. Darling et al⁵⁶  evaluated return to activity using the Zurich guidelines combined with a standardized exertional treadmill test, the Buffalo Concussion Treadmill Test (BCTT). On average, patients took 16 ± 15 days from the time of the concussion until they reported being asymptomatic, with female athletes taking longer than male athletes. However, the time from concussion to physiologic recovery, as defined by passing the BCTT, was longer, and there was a minimal difference between the sexes.⁵⁶  All patients returned to sport within the week after passing the BCTT, which kept them out of sport participation for approximately 1 month. The authors⁵⁶  suggested that a combination of the Zurich guidelines and the BCTT may provide a useful paradigm for making safe return-to-activity decisions.

McCrea et al⁵⁸  evaluated whether a symptom-free waiting period (SFWP) affected clinical outcome and risk of repeat injury among high school and collegiate athletes. Just over 60% of patients had an SFWP, but no differences were found between groups for duration of symptoms, symptom severity score, balance score, or Standardized Assessment of Concussion score at the time of injury or 2 to 3 hours after injury. Patients in the no-SFWP group returned to sport participation approximately 1.2 days before reaching full symptom resolution and 7.1 days earlier than the SFWP group.⁵⁸  No differences were noted for the rate of same-season repeat injury. The authors⁵⁸  proposed that the use of an SFWP neither improved clinical outcomes nor decreased the risk of a same-season repeat concussion.

Despite the significant increase in the number of concussion studies published in medical and scientific journals since the early 2000s, some of the most important decisions regarding concussion management have been based primarily on expert opinion and consensus. In this article, we aimed to systematically review the literature and evaluate the evidence for use of rest, effectiveness of rest, use of and compliance with return-to-activity protocols, and effectiveness of return-to-activity protocols. Our main findings suggest that rest is underused by health care providers, recommendations for cognitive rest do not provide guidance for clinicians during individualized patient care, an initial period of moderate physical and cognitive rest (eg, light physical and mental activity) may improve outcomes during the acute postinjury phase, significant variability exists in the use of assessment tools and compliance with recommended return-to-activity guidelines, and additional research is needed to empirically assess the effectiveness of graded return-to-activity progressions.

The findings regarding the use of rest and compliance with published recommendations also suggest that there is a critical need for education and dissemination of information regarding best practices to community primary care providers for postinjury management and collaboration with concussion specialists and school personnel. Despite the emphasis on cognitive rest and academic adjustments in several consensus and position statements,^1,4,5  authors of most of the included studies identified limited prescription of cognitive rest or academic adjustments for patients after concussion. Further, these studies highlighted deficiencies in the use of multifaceted assessments (graded symptom checklist, neuropsychological testing, and balance assessment) for return-to-play decisions.³⁸⁻⁴²  One reason for these findings may be the isolation of health care providers from other members of the concussion-management team. Effective concussion management requires a team-based approach, in which the family, medical providers, and school personnel work together to manage the patient with concussion.⁵⁹  However, in some settings, there may be a disconnect because physical activity is managed independently of academic activity and communication among team members is limited.⁶⁰  Strategies must be developed to educate all team members and build a collaborative network that includes a standardized approach to evaluate each patient for needed cognitive and physical rest and provide individualized recommendations based on the patient's medical history and clinical presentation. In 1 study,²³  the development of a standardized evaluation and management protocol reduced the variability in the instructions provided to patients regarding return to activity and return to school.

With respect to postinjury management, clinicians need to consider balancing rest and active treatments. Although our findings in this review highlight that too much activity can hinder recovery, they also suggest that strict rest can do the same. The critical message may be that rest is important in the first 1–2 days after concussion but that moderate amounts of physical and cognitive activity, under the direction of a health care provider, may be beneficial in the recovery process. Including moderate-intensity activities may improve the psychological wellbeing of patients, especially in the athletic population.¹²  In patients with persistent symptoms, 1 retrospective chart review²⁷  showed that a 2-week period of rest may be beneficial in reducing symptoms; however, light activity has also been noted to improve recovery in similar patient populations.^14,61  Until additional research has identified the most effective timing and intensity of such early interventions, each clinician must determine the optimal balance of rest and activity for each patient, taking into account the patient's past medical history and current complaints. For athletic trainers, these decisions should be made in conjunction with their directing physician and in collaboration with other concussion-management team members.

Validation of the existing return-to-activity progressions is also needed. Current recommendations suggest waiting until the patient is asymptomatic and has returned to baseline on assessments of cognition and balance before beginning a progressive protocol to return to competition. Yet the literature is unclear as to what constitutes being asymptomatic or whether an earlier introduction to light activity may facilitate recovery.¹²  This latter point may be supported by the work of Darling et al,⁵⁶  who combined the BCTT with the Zurich guidelines and the advice of those suggesting that light activity may be beneficial to recovery.^29−31,36  Furthermore, some evidence supports early active exercise among patients who are otherwise slow to recover¹⁴  or those with postconcussion syndrome.⁶¹ 

In addition to the limited evidence supporting the return-to-activity progression, recommendations have been made to individualize the approach and take into account potential modifying factors, such as age, sex, past medical history, and comorbid conditions (eg, learning disabilities, attention-deficit/hyperactivity disorder).^4,5,62  For pediatric patients, both the Concussion in Sport Group⁴  and the American Academy of Pediatrics²  advocate lengthening this progression by recommending that patients be asymptomatic for a longer period (eg, SFWP) before starting the graded return-to-activity progression. The American Academy of Pediatrics²  recommended a minimum of 5 days to progress through the stages to a full return to activity, with additional time built in for patients having a prolonged recovery or those with a history of concussion.

Although the graded return-to-activity progression has been accepted by most medical professionals as the standard of care for returning athletes to sport participation, this approach has not been substantiated with prospective, randomized controlled comparative-effectiveness trials of this approach on patient outcomes, return to activity, or risk of repeat injury, nor is it considered appropriate for children and adolescents.⁶³  Only 1 research group,⁶⁴⁻⁶⁶ to date, has aimed to develop pediatric-specific protocols for return to activity and return to school. Using scoping methods, DeMatteo et al⁶⁴  aimed to identify and describe the management strategies used by health care providers and the evidence supporting or refuting each strategy. The review of 400 possible articles included information from 10 studies and 3 Web sites; however, none of the guidelines were specific to the younger population. As a result of these findings, follow-up studies using the National Institute of Health and Care Excellence procedures were conducted to develop pediatric-specific return-to-activity⁶⁶  and return-to-school⁶⁵  protocols. These protocols take a more conservative approach, strive to find a balance between allowing the brain to heal and addressing the need to participate in some activity, and emphasize that different pathways are likely required during the recovery process.⁶⁶  After the development of the protocols, pilot testing was conducted with health care providers, and all strongly agreed that using the protocol changed their clinical practice, increased their knowledge, and enhanced their confidence in treating children with concussion.⁶⁶  Similar findings were noted among health care providers who pilot tested the return-to-school protocol: 95% strongly agreed that the protocol gave specific and clear directions for applying the recommendations, 80% indicated that it increased their knowledge, and 85% strongly agreed that their confidence in managing children with concussion improved.⁶⁵  These findings are positive and demonstrate a more evidence-based approach to developing return-to-activity progressions, but future authors needs to empirically evaluate how these approaches affect patient outcomes before protocols such as these are used routinely in clinical practice.

Additional research is warranted to determine the effectiveness of rest, treatment, and returning to activity. The literature and clinical practice would benefit from multisite comparative-effectiveness studies to evaluate the timing, duration, and levels of rest compared with other active treatments. In addition, studies evaluating the effectiveness of the graded return-to-activity progression are needed to provide clinicians with additional information with which to manage patients with concussion. Lastly, studies of knowledge translation for primary care providers are needed. Most of the studies evaluating rest were done in concussion specialty clinics under the supervision of concussion specialists. However, most young patients are evaluated and treated in emergency department or urgent care clinics or by their pediatrician or primary care provider. Efforts are required to disseminate best-practice information regarding rest and return to activity to those providers and evaluate how that knowledge affects patient care.

This systematic review is not without limitations. Because of the descriptive nature of most of the included studies, we did not evaluate the risk of bias. In addition, because of the small numbers of studies included, we were unable to find any studies specifically investigating the effectiveness of the graded return-to-activity progression or differentiating findings based on age or sex. Although we were able to include 40 studies in this review, only 6 were considered level 2 evidence (RCTs or inception cohort studies); of the rest, 29 studies were considered level 3 (nonrandomized cohort or follow-up study), and 5 were level 4 (case control or case series). It is important for clinicians to understand the effect of level of evidence on the ability to assign a strength of recommendation. The findings from studies with lower levels of evidence need to be evaluated cautiously before they are used regularly in clinical practice. Furthermore, the limited number of studies included did not allow us to evaluate the outcomes by setting (emergency department versus clinic) or timing after concussion, which may significantly alter how this information is used for clinical decision making in various environments.

This systematic review presents a descriptive analysis of the use and effectiveness of rest and return to activity after sport-related concussion. Specifically, physical rest and cognitive rest were underused by health care providers (strength of recommendation = B), moderate physical and cognitive rest may facilitate recovery during the initial days after concussion (strength of recommendation = B), significant variability exists in the use of assessment tools and compliance with recommended return-to-activity guidelines (strength of recommendation = B), and little evidence supports the effectiveness of the graded return-to-activity progression (strength of recommendation = D). High-quality, prospective studies evaluating the influence of rest, early light exercise, and other treatment options are needed to provide an evidence-based road map for managing patients with sport-related concussion. Until additional research has been published, it is prudent for clinicians to approach concussion management and return to activity in a conservative manner that evaluates the clinical presentation and the needs of each patient.

We thank Casey Erickson, ATC, and Krista Hixson, ATC, for their assistance in organizing the search results for this manuscript.