The effect of footwear and sports-surface on dynamic neurological screening for sport-related concussion

1. Introduction 

Concussion and mild traumatic brain injury (mTBI) are common injuries in contact and collision sports,1, 2, 3 however, it has been noted that many concussions go undetected and unreported.4, 5, 6 Unreported concussions are serious because they increase the likelihood of athletes continuing to practice and play, placing themselves at risk of a subsequent and possibly more serious injury. Valid and reliable screening tools are therefore essential in assisting the sports medicine professional in recognising a concussion.

The International Concussion in Sport Group is developing a standardised screening and assessment tool for the sideline screening of sport concussion.7 The Sport Concussion Assessment Tool (SCAT) was designed to assess a number of specific concussion indicators including physical signs and symptoms; memory and cognitive performance; and a neurological screen of motor performance. The updated SCAT2 recommends a qualitative measurement of gait performance; however, there is no specific assessment protocol associated with the item and it is thus open to differing interpretations.

This issue has been recently addressed by the introduction of a timed Tandem Gait (TG) task8 that evaluates locomotion, dynamic balance and lower limb coordination. The test requires the individual to walk in a heel–toe manner along a 3-m line, turn and return to the starting position as quickly as possible without losing their balance. The TG task has been shown to be reliable in young non-concussed barefoot individuals in a controlled environment and norms have been developed.9

Testing environment,10 footwear,11 and support surface12, 13 have all been shown to affect the measurement of balance and gait performance. It has been demonstrated that balance testing using the Balance Error Scoring System (BESS), which is frequently used in the screening and assessment of a concussion,7, 14 is affected by the environment in which it is administered. Athletes are often screened in settings such as; emergency rooms, changing rooms or on the sidelines, and often the sports-surface and footwear are specific to the sport. Improved scores were obtained when the BESS was administered in a locker room environment in comparison to when it was administered on the sideline during sports participation.10 This finding has implications for the interpretation of performance scores in a potentially concussed athlete who may be assessed serially in different environments and conditions, and potentially by different health care providers. The types of footwear and surface have been shown to affect performance during gait assessment, with participants walking faster with shoes11 and slower when walking on irregular surfaces.12

Therefore, it is vital to understand the difference in performance of a potential concussion screening task such as TG when measured under different environmental conditions. The purpose of this study was to investigate the effect of footwear and sporting surface on the time to perform the standardised TG motor performance task.

2. Method 

3. Results 

Thirty-six participants were recruited from each of three sporting environments (groups). No participants failed any trial via our criteria, however, one participant in the artificial turf group produced atypically slow times (>35s), which indicated that they were most probably not following the instructional set, and their data were removed from further analyses. The characteristics of the participants are presented in Table 1. The three groups were comparable (p>0.05) with respect to age, weight and BMI while the artificial turf group was shorter (p=0.03) than the other two groups (F2,105=4.76, p=0.011) which likely reflects the higher ratio of females in this group. In contrast to the turf participants, the grass and court groups were predominantly male. The mean stud/cleat length of the football boots worn on grass was 14.6±3.2mm and the mean ground hardness was 56 (±9.6)N/cm2 with these results included primarily to allow study replication. Inter-rater reliability of TG timed measurement for combinations of raters resulted in ICC (2, 1) values ranging from of 0.96 (C.I. 0.80–0.99) to 0.99 (C.I. 0.98–0.99).

Table 1. Characteristics of participants grouped by primary sports-surface.

	Sports-surface	Age (yrs)	Sex	Height (cm)	Weight (kg)	BMI
		Mean±SD	Male:female	Mean±SD	Mean±SD	Mean±SD
	Grass (N=36)	20.17±1.34	30:6	176.64±8.55	74.86±11.26	23.91±2.53
	Court (N=36)	20.64±2.19	23:13	175.90±8.87	73.72±12.14	23.84±2.68
	Artificial turf (N=35)	20.29±3.27	18:17	170.90±9.61	69.69±11.83	23.72±2.56
	Group difference	F2,104=0.337 p=0.687	χ2=9.536 p=0.008	F2,104=4.24 p=0.017	F2,104=1.883 p=0.157	F2,104=0.045 p=0.965

The TG task times for the various conditions are presented in Table 2, Table 3. In the 3×2 ANOVA there was a significant main effect by condition (F2,138=26.31, p=0.001) with all 3 groups faster with footwear than barefoot (Table 2). This difference was greatest in the artificial turf and court athletes resulting in a significant group×condition interaction (F2,104=3.35. p=0.039). In the 2×2 ANOVA there was no difference between the two barefoot conditions within each environment (F1,69=0.10, p=0.751) (Table 3.). Examination of the two barefoot conditions for the grass and artificial turf groups showed a significant interaction (F1,69=15.66, p=0.000) with the artificial turf athletes slower. It is noted that despite the standard deviations of the mean values being relatively tight, the distributions (range) do overlap between all conditions which reflects some variability in task performance that should be taken into consideration when interpreting the results.

Table 2. Tandem Gait time(s) for sports-surface and footwear conditions.

	Sports-surface	Test condition
		Footwear	Barefoot
	Grass (N=36)
	Mean±SD	12.80±2.27	13.90±3.08
	Range	8.25–17.47	8.60–20.89
	Court (N=36)
	Mean±SD	13.89±2.87	16.18±3.18
	Range	8.45–20.46	11.15–23.10
	Artificial turf (N=35)
	Mean±SD	14.48±2.68	16.30±2.55
	Range	9.52–22.26	10.19–22.88

Table 3. Tandem Gait time(s) for sports-surface groups during barefoot conditions.

	Sports-surface group	Test condition
		Barefoot on sports-surface	Barefoot on concrete
	Grass (N=36)
	Mean±SD	13.90±3.08	13.77±2.75
	Range	8.60–20.89	8.39–20.79
	Artificial turf (N=35)
	Mean±SD	16.30±2.55	16.58±3.41
	Range	10.19–22.88	10.48–26.31

4. Discussion 

The results demonstrated that performance on the timed TG task varied significantly with the nature of the sporting group or surface and whether participants were barefoot or shod. Footwear (boots or sports shoes) resulted in a faster performance than when the TG task was performed barefoot. Likewise, times varied between each sporting environment which might suggest that the sporting surface or the type of athletes played a role in TG performance.

The reason for the significantly faster performance on all testing surfaces while wearing footwear is speculative. The logical and practical explanation for decreased speed when walking heel–toe barefoot could be that the participants were more careful to avoid standing on their toes and therefore walked more carefully (slowly). Likewise, the higher level of sensory feedback while performing barefoot16 could have created a more careful gait pattern in fear of getting hurt. Additionally, shoes might provide a larger base of support and could afford additional stability during dynamic balance and coordination tasks. The enhanced performance while wearing footwear has been demonstrated by others during traditional gait and balance tasks,11, 17 illustrating that the footwear provides some form of protective function and/or confidence to the user, allowing them increased speed and accuracy. The lack of difference between the barefoot conditions on grass/artificial turf and concrete suggests that the change in surface was not a distinguishing factor and justifies our decision not to test an alternative firm surface (concrete) within the hardwood court environment. The fact that barefoot times did not significantly change between firm and more compliant surfaces also suggests that the physical, sporting and psychological characteristics of each sports group were more likely responsible for the difference in performance observed between environments.

Previous data on TG in young persons has been generated while walking barefoot or with sports shoes in a clinical laboratory environment.8, 9 The TG times reported in this present study are slower than those previously reported when walking with8 or without shoes,9 despite the firm court and concrete surfaces approximating the laboratory (vinyl on concrete) surface of the former studies. The differences observed might therefore reflect the sports environment, with increased participant distraction due to other competing and concurrent activities, noise and varying climatic conditions. This finding is in keeping with a recent study which measured static balance using the BESS in different environments10 and demonstrated that sideline testing resulted in decreased performance compared to a controlled clinical environment (locker room). Our data adds to the argument that repeated measurement of standardised tests must also take into account when and where the measures are obtained. As the TG task has the potential to be used as both a preseason assessment and for the monitoring of recovery in SRC, it is important that the conditions surrounding each assessment are well documented to assist in the interpretation of the subsequent scores.

A difference in TG performance between sporting groups was also identified in this study. The participants were persons who currently were playing sport in each of the three environments and the ages, weights and BMI were comparative between the groups. While gender differences were not balanced in two (grass and hardwood court) of the three groups and might have affected the results, the gender ratio is typical of the sports studied (rugby, football, basketball) and the higher percentage of males who sustain a concussion.18 In a preliminary study by the authors,8 females were shown to be significantly slower during TG than males; however, two recent larger studies reported no gender effect with TG.9, 19 An alternative reason for the differences between groups might be the motivation and competitiveness of the participants. Anecdotally, the football/grass group, who performed best across all conditions, consisted of associated club members who brought an element of competition to the testing arena (team practice). While this factor was difficult to minimise in the open environment where the testing took place it is unlikely to be present during screening of an injured/concussed athlete. It must also be noted that a possible limitation to the reliability of results exists due the use of a manual stopwatch.

5. Conclusion 

The neurological assessment of sports concussion with standardised motor performance tests is common; therefore it is vital that the effect of the footwear and sporting surface on these tests is understood. This study investigated the effect of footwear and sporting surface on the timed TG task, a dynamic measure of speed, coordination and balance. The results demonstrated that the time for a defined TG task depends on whether footwear is being worn and we propose that sporting surface and motivation might also be factors in performance outcome. In addition, there were relatively large variations in task performance within conditions. Therefore it is recommended that baseline assessment and subsequent screening of dynamic motor tasks are performed with the same footwear and, if possible, in the same environment for each sports participant.

Practical implications