Normative Performance on the Balance Error Scoring System by Youth, High School, and Collegiate Athletes

An estimated 30 to 45 million children participate in nonscholastic athletic programs annually in the United States.¹ Concussion is a common injury among youth, adolescent, and collegiate athletes,² with an estimated 1.6 to 3.8 million sport-related concussions occurring annually in the United States.³ Whereas epidemiologic and observational studies of concussion have been prevalent among high school and collegiate athletes,⁴ specific methods to evaluate concussive injuries among youth athletes are lacking, leaving this youngest, and potentially most vulnerable population, underrepresented and unstudied.⁵ Concussive injuries induce a complex pathophysiologic response, resulting in a myriad of symptoms and short-lived neurologic impairment involving the cognitive, motor, visual, and vestibular domains of function.⁶ Postural instability is a hallmark of acute concussion. Using subjective clinical evaluation, researchers⁷ have proposed that postural-stability declines may resolve in high school and collegiate athletes within 3 to 10 days postinjury; however, investigators⁸ using more sophisticated biomechanical measures of balance have indicated that deficits may persist for weeks or months. Declines in postural stability after concussion are caused by disruptions in the integration of sensory information from the visual, vestibular, and somatosensory inputs and resultant inadequate motor response.⁹ The incidence of postural instability and its importance in motor-control processes have led to the recommendation that balance assessment be a cornerstone of concussion evaluation from healthy baseline to return-to-play decision making for high school and collegiate athletes.¹⁰ A gap in managing youth concussion is the lack of understanding about how children perform on standard clinical balance assessments designed for concussion assessment.

The Balance Error Scoring System (BESS) is a clinical measure that was originally designed to characterize postural stability in high school and collegiate athletes in order to detect differences between concussed and healthy athletes up to 3 to 5 days postinjury.¹¹ In its entirety, the BESS involves testing individuals in 3 stances (double legged, single legged, and tandem) on firm and foam surfaces. After the 3rd International Conference on Concussion in Sport held in 2008, a modified version of the BESS, in which the foam surface is omitted, was adopted for use in the Sport Concussion Assessment Tool 2 (SCAT2) and became the most commonly used sideline measure of balance.¹² Whereas subsequent consensus statements on concussion in sport have acknowledged that the modified BESS is not well studied and using the foam surface improves its sensitivity,¹³ both updated versions of the SCAT (SCAT3 and SCAT5) continue to use only the stances on a firm surface.^6,14 Nonetheless, given the ceiling effect evident with the modified BESS and the presumed prevalence of in-clinic use of the entire BESS, it is critical to understand the relationship among healthy children's, adolescents', and young adults' BESS performances.

In the absence of individual healthy baseline values, normative data are compared with postinjury balance-performance values to determine the residual balance impairment or degree of recovery. Whereas normative BESS data are available for high school and collegiate athletes,^15–17 studies^17–19 quantifying BESS performance in youth athletes, particularly those less than age 14 years, have produced results that were limited and conflicting. Successfully managing youth concussion requires age- and sex-specific normative values due to the rapidly developing central nervous system processes underlying postural stability.

The need for age-specific normative values for the BESS is clear, as authors^20,21 of biomechanical studies have observed developmental and sex differences in postural control throughout childhood, adolescence, and adulthood. Females generally exhibit less postural sway than males up to approximately age 11 years, suggesting more rapid maturation of the integration of afferent information to maintain a steady posture.²⁰ However, clinical balance assessments, such as the BESS, may not have sufficient resolution to detect the subtle differences in the maturation of balance control between males and females that are evident with biomechanical assessments. Olson¹⁸ reported that sex and age did not affect BESS scores when comparing 11- to 13-year-old individuals with 17- to 18-year-old individuals. Others^19,22 have reported that sex, but not age, influenced BESS scores. Whereas normative BESS performance data have been obtained in smaller cohorts of youth,^18,22 high school,^18,23 and collegiate^23,24 athletes, to date no cross-sectional data exist that systematically characterize BESS performance across a wide range of ages among males and females. The lack of cross-sectional data reflects a fundamental gap that complicates the effective management of concussion in the pediatric athlete. Therefore, the purpose of our study was to determine age- and sex-specific normative values for the BESS among a large cohort of healthy youth, high school, and collegiate athletes.

METHODS

Sample

We performed a retrospective chart review of 6762 athletes (4600 males, 2162 females; age range = 5–23 years) who completed baseline testing during the 2013–2014 athletic seasons. Participant demographics are provided in Table 1. Balance assessments were completed as part of routine preseason baseline testing at sport clubs, high schools, and colleges conducted by personnel from the Cleveland Clinic's Concussion Center. All participants were neurologically healthy, and none presented with active musculoskeletal impairments that affected postural stability or precluded their participation in sport. Athletes recovering from musculoskeletal injuries, such as sprains, strains, and fractures, and those rehabilitating after surgery were excluded from the sample. The Cleveland Clinic Institutional Review Board approved this retrospective analysis.

Table 1 Sample Demographics

View Fullscreen

Data Collection

Before data collection, experienced raters on the study staff (S.J.O., S.M.L., J.C., J.L.A.) performed hands-on training for each athletic trainer to ensure optimal consistency of administration. The BESS was then administered by 48 certified athletic trainers using standardized testing procedures.²⁵ Each participant completed the six 20-second trials according to the BESS protocol: double-legged, single-legged, and tandem stances on firm and foam (Balance Pad; Airex AG, Sins, Switzerland) surfaces in a quiet room (Figure 1). Participants placed their hands on their iliac crests and were instructed to stand as still as possible and close their eyes. When they closed their eyes, the 20-second trial was initiated. Errors were recorded using standardized procedures, and the maximum score per stance was 10.⁶ In addition to the BESS, participants completed a neurologic test battery of iPad (Apple, Inc, Cupertino, CA) applications that provided quantitative assessments of cognitive, fine-motor, and visual performance, which are beyond the scope of this paper.

View Figure

• Download as Powerpoint
• Download Figure

RESULTS

The median (minimum to maximum) scores for the total BESS, firm BESS, and foam BESS for the entire sample (N = 6762) were 13 (0–50), 3 (0–22), and 10 (0–30), respectively. The greatest number of errors occurred during single-legged stance on a foam surface (7 [0–10]), followed by tandem stance on a foam surface (3 [0–10]) and single-legged stance on a firm surface (2 [0–10]). During the double-legged stance on a firm surface, a ceiling effect was evident (0 [0–6]), which was similar to that for the double-legged stance on a foam surface (0 [0–10]) and tandem stance on a firm surface (0 [0–10]). These error-count trends were consistent across sex and age stratifications as depicted in Figures 2 and 3.

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Young Males Exhibited the Greatest BESS Errors

The multivariate comparisons of total, firm, and foam BESS scores revealed an age effect across all cohorts, specifically youth (5–13 years), high school (14–18 years), and collegiate (19–23 years) athletes, and across all 6 balance stances (Pillai trace = 0.022, F_12,13504 = 12.44, P < .001; η²_p = 0.011; medium effect size). Table 2 provides between-subjects effects. When we accounted for sex, youth males committed 31%, 59%, and 20% more errors than high school males and 51%, 98%, and 35% more errors than collegiate males for the total BESS, firm BESS, and foam BESS, respectively (Figure 2 and Table 3). In general, less variability and slightly different trends were observed among females. Youth females exhibited 4% and 8% lower error rates than high school females for the total BESS and foam BESS, respectively, and performed similarly to collegiate females on the foam BESS. The inverse was observed for performance on the firm BESS, with 9% and 20% more errors committed by youth females than their high school and collegiate female counterparts, respectively.

Table 2 Sex, Age, and Sex × Age Differences for Balance Error Scoring System (BESS) Conditions (2-Way Analysis of Variance)

View Fullscreen

Table 3 Balance Error Scoring System (BESS) Scores Across Stances Stratified by Age and Age × Sex

View Fullscreen

When we further stratified youth athletes into smaller age brackets, a performance gap was evident between 5- to 9-year-old males and all other athletes for the total, firm, and foam BESS. This trend was not observed among the youngest female cohort, as 5- to 9-year-old females performed as well as or better than their older female counterparts. Post hoc univariate ANOVAs demonstrated differences between 5- to 9-year-old and 10- to 13-year-old athletes during the total (F_3,6756 = 33.56, P < .001), firm (F_3,6756 = 45.59, P < .001), and foam (F_3,6756 = 14.43, P = .02) BESS and during the single-legged stance on a firm surface (F_3,6756 = 37.06, P < .001) and the tandem stance on firm (F_3,6756 = 18.32, P < .001) and foam (F_3,6756 = 8.09, P = .02) surfaces.

Percentile Scores Based on Age and Sex Cohorts

Percentile scores for the total, firm, and foam BESS and each BESS condition are presented according to age and sex cohorts (Table 4). Given the performance gap observed in the youth male cohort, 5- to 13-year-old athletes were subdivided into 2 age brackets: 5- to 9-year-old and 10- to 13-year-old athletes. Whereas this performance gap was not evident in our sample of youth females, percentile scores are still reported separately for 5- to 9-year-old and 10- to 13-year-old athletes, as variability within the youngest cohort was higher than for the high school and collegiate athletes.

Table 4 Age- and Sex-Specific Percentile Scores of the Balance Error Scoring System (BESS) Continued on Next Page

View Fullscreen

DISCUSSION

Evaluating postural stability is essential for effective management of athletes with concussion. The BESS is the most common clinical balance assessment used to determine deficits in postural stability after a concussion. Interpreting the BESS is critical to informing clinical decisions about the concussion diagnosis, recovery, and return to play. However, the absence of age- and sex-specific normative values complicates its use in pediatric patients. In the absence of baseline assessments, our results underscore the need to use age- and sex-specific norms for youth, high school, and collegiate athletes. They indicated that age and sex affected BESS performance in each defined cohort. In general, (1) BESS performance in collegiate athletes was superior to that in high school followed by youth athletes; (2) BESS performance in females was superior to that in males, particularly in the youth and high school populations; and (3) ceiling or floor effects were evident in 3 of the 6 BESS conditions (ie, double-legged stance on firm and foam surfaces and single-legged stance on a foam surface).

Improved BESS performance as a function of age is likely due to the maturation of 2 physiological processes: the integration of sensory systems (ie, visual, vestibular, and somatosensory) and the development of automatic motor processes.³¹ Motor processes mature early in childhood (at 3–4 years of age), whereas sensory integration does not reach adult levels until 14 to 16 years of age.^20,21 Therefore, maturation of the sensorimotor systems responsible for maintaining balance is likely complete in older adolescents. The age-related reduction in BESS error-count variability is consistent with data²¹ demonstrating that postural stability matures with age; youths committed more errors than high school and collegiate athletes on the total, firm, and foam BESS and across all 6 BESS conditions. Further dividing the broader age cohorts (youth, high school, and collegiate) into smaller age ranges demonstrated the effect of maturation, as the youngest males (5–9 years of age) committed more total BESS errors than did older cohorts. Heterogeneity in BESS errors was observed in the youngest bracket (5–9 years of age), likely due to variability in the rate of development and in the maturation of the motor control necessary for balance maintenance. The neuronal domains of postural control challenged by the BESS appeared to stabilize in our sample in the early teen years, as evidenced by the reduction in total errors and error-score variability in the high school and collegiate cohorts. The normative values across age groups in our dataset were consistent with previous findings^15,19 indicating improved postural stability as children and adolescents age.

Analyzing normal performance on a stance-by-stance basis serves multiple purposes. Namely, the degree to which each stance taxes the balance system can be compared. One would expect that youths would demonstrate worse performance for all stances, as their control of postural stability is still maturing. However, whereas age-specific differences were statistically significant across all stances, differences in performance on both double-legged stances (foam and firm surfaces) were likely not clinically meaningful, as the mean error scores across the 3 age cohorts ranged from 0.02 to 0.04 on a firm surface and 0.11 to 0.34 on a foam surface. All ages included in the sample likely had sufficient sensory integration and motor control to not exhibit “errors” or discrete losses of balance in these relatively simple stances. In fact, 99.5% and 96.6% of all participants incurred either zero errors or 1 error on the double-legged stance on firm and foam surfaces, respectively.

Sex differences have also been reported^20,21 in the maintenance of postural stability and vary with age due to sensorimotor function and anthropometric and psychological factors. Vestibular system maturity has been shown to occur earlier in females than in males.²¹ Postural stability also depends on anthropometric properties that vary between sexes throughout childhood and adolescence, including body mass and center of mass (ie, a point in space that represents the weighted average of each body segment in space).³² It is inversely proportional to the center of mass, which is lower among females due to a smaller waist-to-hip ratio, greater maximal thigh girth, and narrower shoulder width.³³ Researchers have established that standing on the foam pad decreases postural stability by disrupting the reliability of the somatosensory input³⁴; however, because males are likely to have a larger mass than females, the increased deflection of the balance pad could theoretically provide an advantage in the BESS foam conditions, particularly for those weighing more than 90 kg.³⁵ Lastly, psychological factors, including motivation and attentiveness, may explain sex differences among age cohorts.²¹ Males were less attentive and more agitated during postural-stability tasks, which could result in worse performance during balance testing.^20,21 Our normative values across sexes are consistent with previous findings on the BESS,^22,24,36 except for those of Olson,¹⁸ who did not find sex differences between youth and high school athletes. This absence of a sex effect may have been due to the smaller age ranges and smaller sample sizes of males and females within each group than in our study.

Youths and adolescents exhibit continual and often rapid physical growth relative to young adults. During this time of development, sex is an important factor that affects the development of postural control.³⁷ Our normative dataset indicated that the developmental factors affecting postural stability (ie, maturation of sensory information, anthropometrics, and psychological factors) did not affect performance on the BESS in either sex after the age of 14 years. However, differences were evident in male and female youths, signifying the need to further stratify the youngest cohort into 5- to 9-year-old and 10- to 13-year-old age brackets when reporting normative values. Therefore, BESS error scores are presented as a function of percentiles of the normative dataset (see Table 4). Whereas we did not examine the psychometric properties of the BESS, Finnoff et al³⁸ reported relatively low interrater and intrarater reliability for the BESS scores and resultant interrater and intrarater minimal detectable changes of 9.4 and 7.3 points, respectively. In addition, researchers^39–41 have found evidence of a practice or learning effect from previous exposure to the BESS or retention of improvement over time (or both) in youth, high school, and collegiate athletes up to 60 days after baseline assessment, potentially resulting in premature clearance before full recovery. As such, minimal detectable change values and possible practice or learning effects highlight the subjectivity and limitations of the BESS and should be considered when interpreting results as a whole and when determining clinically relevant differences in BESS performance.

Although the BESS is portable and quick to administer, our normative data indicated that it has floor and ceiling effects. During both the double-legged stances on firm and foam surfaces and the tandem stance on a firm surface, BESS errors were infrequent, with 99.5%, 96.6%, and 76.5% of all athletes producing no errors or 1 error, respectively. Conversely, during the single-legged stance on a foam surface, 53.2% of all athletes committed 8 to 10 (maximum number) errors, and 20% of all athletes committed 10 errors. The large proportion of nonconcussed athletes committing 8 to 10 errors at baseline during the single-legged stance on a foam surface raises concerns over the utility of this BESS condition in identifying impaired balance after a suspected concussion. A minimum or maximum error score in a given condition threatens the sensitivity of the measure and limits its resolution in detecting balance differences that may otherwise be present due to age, sex, or injury status.

The modified version of the BESS (without the use of a foam pad) was recommended for assessing postural stability by the 3rd, 4th, and 5th International Conferences on Concussion in Sport and has been incorporated into the widely used SCAT (versions SCAT2, SCAT3, and SCAT5).^6,12,14 In discussing the utility of the modified BESS and its use in the SCAT3, Guskiewicz et al¹³ acknowledged that most of the research on the BESS has included firm and foam surfaces with variable interrater and intrarater reliability and that additional study was required to determine if the firm conditions alone were adequate to evaluate balance. For example, Finnoff et al³⁸ reported the lowest interrater reliability (intraclass correlation coefficient = 0.44) for the tandem stance on a firm surface relative to all other BESS stances, but they could not determine the intraclass correlation coefficient for the double-legged stance on a firm surface due to a lack of between-subjects variability. Our results revealed floor effects in 2 of the 3 stances, and the mean score for the 3 firm stances combined was only 3.9, or 27% of the total BESS score of 14.2 ± 6.9. Based on previous literature and our results, the modified BESS alone may be insufficient for detecting deficits in postural stability or adequately informing the appropriate timing of the return-to-play phase of concussion management.

Whereas our normative data were an attempt to improve the utility of the BESS by stratifying according to sex and age (youth, high school, and collegiate) cohorts, BESS performance should always be interpreted in the context of all clinical information, including the athlete's medical history, to avoid clinical decision-making errors when preinjury baseline data are unavailable.

CONCLUSIONS

Accurately assessing and interpreting postural stability is critical for effectively managing athletes with concussion. Given the developmental differences between age and sex cohorts, using the normative reference values we reported will enable a more reliable interpretation of postural-stability assessment for clinical decision making. Researchers should focus on improving the interpretation of the BESS by determining the minimal clinically important difference in scores when comparing healthy and injured populations and on developing biomechanical measures to reduce the subjectivity associated with balance testing.