ASSESSING STATIC AND DYNAMIC POSTURAL CONTROL IN A HEALTHY POPULATION

Abstract

Static postural control testing is often conducted by clinicians and athletic trainers for use with athletes who have sustained a concussion. Dynamic postural control involves the body’s response to perturbation of the center of mass and may offer additional insight that static testing cannot capture. The objective of this study was to assess the reliability and feasibility of a balance protocol consisting of both static and dynamic postural control assessments with a healthy, adult population. Subjects stood in both unipedal and bipedal stances on a force plate to capture quantitative data regarding the center of pressure over time. Further, subjects completed the Balance Error Scoring System (BESS), a static measure, and a modified version of the Star Excursion Balance Test (SEBT), a dynamic measure. Reliability with the BESS was limited, while moderate to strong reliability was obtained for the modified SEBT. Unipedal stances were associated with a greater variance than bipedal stances for both the BESS and force plate protocol. These assessments will be applied within a pediatric populations to determine the validity of their use. Further postural control research is necessary to determine the most viable assessments for use within an active, pediatric population.

INTRODUCTION

After mild traumatic brain injury (mTBI), it is common for postural control deficits to be observed [1–3]. Many post-concussion assessments now include postural control tests as an evaluative tool to determine patient health [4–7]. Postural control represents the ability of a person to maintain balance naturally and when exposed to perturbation [8]. Postural control can be defined by assessing static and dynamic balance. Static balance involves an individual establishing a stable base and attempting to minimize movement while holding the particular posture. Dynamic balance, on the other hand, refers to the introduction of perturbations to this stable base of support. It can be assessed by having subjects establish a base of support and then requiring some level of movement away from that equilibrium. Static balance has been most commonly assessed in post-concussion situations, though dynamic balance assessments are gaining favor as they may involve movements similar to those experienced while playing sports [9–11].

Static balance is most commonly assessed using the Balance Error Scoring System (BESS) or force plates. The BESS is an easily administered, static balance assessment for sideline use in instances of suspected concussion that asks individuals to hold different static postures while an evaluator assesses deviations from this desired posture [12, 13]. Instrumented force plates are used to quantitatively track the center of pressure (COP) over time during a static stance.

Dynamic balance assessments are necessarily more involved than are static assessments, and have seen less use [14]. One of the most commonly employed assessments is the Star Excursion Balance Test (SEBT), which tasks individuals with maintaining balance with one foot while reaching out in prescribed directions with the other foot [9]. By more closely aligning concussion testing assessments with physical activity, it is hypothesized that the tools will be more relevant. The SEBT is traditionally used to assess ankle instability and risk of lower extremity injury, but it is reasonable to expect this dynamic postural control assessment to differentiate concussed athletes as well.

As pediatric populations are still developing postural control, there is a lack of research investigating either static or dynamic postural control within a youth population [8, 10, 14, 15]. Before assessing postural control with this population, the protocol must first be investigated in a population with fully developed balance. The objective of this study was to assess the reliability and feasibility of a balance protocol consisting of both static and dynamic postural control assessments with a healthy, adult population. Static assessments included the BESS and a force plate protocol while dynamic postural control was assessed using a modified version of the SEBT. Upon successful completion of this study, the protocol would then be studied in a cohort of youth football players.

METHODS

Ten healthy, male subjects completed the balance assessments in this study approved by the Virginia Tech Institutional Review Board (IRB-17–1023). These subjects had an average age of 22.2 ± 1.5 years, an average body mass of 87.6 ± 12.2 kg, and an average height of 1.78 ± 0.06 m. Subjects completed two rounds of testing, with the sessions separated by a week. A medical history questionnaire was completed by each subject to assess relevant injuries, specifically concussion history and lower extremity injuries [16]. Testing was conducted by trained lab personnel and consisted of administration of the BESS, a force plate protocol, and a modified version of the SEBT.

The BESS consists of subjects holding specific stances for 20 second durations with their hands on their hips and eyes closed. The test stances (double leg, single leg, and tandem) are carried out on two surfaces (flat ground and a foam pad [Airex Balance Pad 81000, Power Systems, Knoxville, TN]). Scoring for the BESS consists of the researcher counting the number of participant errors within each stance [17]. Thus, lower BESS scores are associated with better postural control.

For the force plate assessment, subjects completed four trials of 30 seconds each: bipedal eyes open, bipedal eyes closed, unipedal eyes open, and unipedal eyes closed. The unipedal stances were added to increase the difficulty of the static postural control tasks. An IsoBALANCE^®2.0 (IsoTechnology, Australia) operating at a frequency of approximately 10 Hz collected COP data in the medial-lateral (ML) and anterior-posterior (AP) axes. All subjects were aligned similarly, with their feet situated along lines on the force plate and the COP centered on the force plate.

Five measures of spatial variability within the COP trajectory were computed: AP and ML sway, path length, maximum path velocity, and 95% COP area. AP and ML Sway represent the standard deviation of the COP trajectory along the respective axis of motion. Path length represents a summation of overall distance travelled by the COP during the 30 second trial, while maximum path velocity assesses the most rapid change in direction during the 30 second trial. Lastly, the 95% COP area represents the area of an ellipse that would capture the mean COP in 95% of samples. Four measures of temporal variability, or entropy, within the COP trajectories were also calculated for each trial: AP and ML Entropy, Renyi Entropy, and Shannon Entropy. Trials with lower levels of entropy, or variability, are considered representative of better static postural control. Background on the determination of these metrics may be found elsewhere [18–20].

The modified version of the SEBT has been previously presented and utilizes 3 of the 8 reaching directions from the original SEBT: anterior, posterior lateral, and posterior medial [21]. Subjects stand on one foot with hands on hips and reach out with the most distal portion of the other foot in one of the prescribed directions before returning to a bipedal stance. Subjects must not transfer weight away from their support foot and cannot lose balance, otherwise the trial must be repeated. Three trials with each leg are completed for the 3 directions. The three trials for each leg-direction combination are averaged and then normalized to the subject’s limb length to account for differences in reach associated with varying anthropometry.

The force plate frequency of approximately 10 Hz is low compared to what is traditionally done (100 Hz). Previous research has shown that more than 95% of the signal power for bipedal stances is below 2 Hz and less than 1% of power is above 10 Hz, which would make a 10 Hz sampling frequency sufficient for measuring static balance [22, 23]. Research with static, unipedal stances has shown that the COP trajectory remains low frequency (95% < 2 Hz) while resulting in greater variability than bipedal stances [24, 25].

Postural control assessment test-retest reliability was assessed using intraclass correlation coefficients (ICC). ICCs with 95% confidence intervals were computed for each force plate metric for the four trials and each stance of the modified SEBT and the BESS. A reliable assessment would be one in which a healthy subject’s score/performance would not vary over time. Higher correlation coefficients are thus associated with reliable assessments.

RESULTS

The overall number of errors measured on the BESS decreased between the two sessions for almost all subjects (Figure 1). These improvements on the BESS were mostly noted for the one foot and tandem stances on a flat surface (Figure 2). Reliability for both two foot stances could not be calculated, as there were no errors observed in these stances. The one foot and tandem stances were associated with low reliability, though the total number of errors from the three foam stances was associated with moderate reliability (Table 1).

Figure 1.

Subjects experienced fewer errors with the BESS during the second session than in the first session (left). Composite reach was consistent between sessions and limbs on average, with subjects reaching 75–85% of limb length (right).

Figure 2.

Matched differences between sessions for the BESS and the modified SEBT. In general, subjects performed better on the BESS during the 2^nd session (left). Individual performance varied between sessions for the modified SEBT, with some subjects reaching further and others not reaching as far (right).

Table 1.

ICC values for each stance of the BESS and Modified SEBT. ICC values are reported as calculated, with 95% confidence intervals shown in parentheses.

			ICC
BESS	Flat	Two Foot	-
		One Foot	0.50 (0.00–0.84)
		Tandem	0.33 (0.00–0.75)
		Total	0.52 (0.00–0.85)
	Foam	Two Foot	-
		One Foot	0.52 (0.00–0.86)
		Tandem	0.54 (0.00–0.86)
		Total	0.81 (0.40–0.95)
Modified SEBT	Left Leg	Anterior	0.53 (0.00–0.86)
		Posterior Medial	0.93 (0.65–0.98)
		Posterior Lateral	0.77 (0.30–0.94)
		Composite	0.96 (0.86–0.99)
	Right Leg	Anterior	0.72 (0.19–0.92)
		Posterior Medial	0.87 (0.58–0.97)
		Posterior Lateral	0.87 (0.59–0.97)
		Composite	0.89 (0.62–0.97)

Performance on the modified SEBT did not vary between the two sessions nor between left and right stances (Figure 1). Reaches while standing on the right foot were associated with greater variation between the two sessions (Figure 2). Reaches in the posterior medial and posterior lateral directions were associated with high levels of test-retest reliability (ICC > 0.75). The composite values for each leg were associated with higher levels of reliability than any individual reach direction (Table 1).

Renyi and Shannon entropy measurements were consistent across both sessions and all test conditions for the force plate (Tables 2 and 3). ML Entropy was observed to increase for unipedal stances compared to bipedal stances. The spatially varying force plate metrics were increased for unipedal stances as well. The unipedal stance with eyes closed was associated with spatially varying parameters that were, on average, at least 2.5 times higher than that observed for the unipedal stance with eyes open. The average matched differences in force plate metrics between the two sessions were close to zero, though individual performance varied.

Table 2.

Summary of Eyes Open Testing. Session 1 and 2 data are reported as mean (standard deviation). The unipedal stance with eyes open was more difficult for subjects than bipedal standing, as evidenced by the spatial varying parameters.

	Static Bipedal			Static Unipedal
	Session 1	Session 2	Matched Difference	Session 1	Session 2	Matched Difference
AP Sway (cm)	0.47 (0.44)	0.42(0.15)	−0.02 (0.16)	0.68 (0.14)	0.61 (0.12)	−0.03 (0.07)
ML Sway (cm)	0.23(0.12)	0.22 (0.09)	0.00 (0.05)	0.55 (0.11)	0.62 (0.15)	0.03 (0.07)
Path Length (cm)	17.18 (11.12)	14.36 (5.12)	−1.11 (3.80)	65.86 (24.90)	60.04 (11.54)	−2.29 (9.07)
Max Path Velocity (cm/s)	2.41 (1.09)	2.06 (0.36)	−0.14 (0.37)	7.17 (2.15)	6.35 (1.48)	−0.32 (0.89)
95% Ellipse Area (cm2)	2.75 (4.60)	1.60 (1.05)	−0.18 (0.66)	7.15 (2.63)	7.21 (2.87)	0.01 (0.55)
AP Entropy	0.37(0.12)	0.35 (0.06)	−0.01 (0.10)	0.65 (0.05)	0.70 (0.08)	0.05 (0.06)
ML Entropy	0.30 (0.12)	0.25 (0.08)	−0.05 (0.09)	0.72 (0.06)	0.68 (0.06)	−0.04 (0.09)
Renyi Entropy	3.56 (0.28)	3.62 (0.25)	0.06 (0.39)	3.81 (0.17)	3.74 (0.23)	−0.07 (0.30)
Shannon Entropy	2.58 (0.13)	2.48 (0.14)	−0.10(0.19)	2.73 (0.04)	2.74 (0.06)	0.01 (0.04)

Table 3.

Summary of Eyes Closed Testing. Session 1 and 2 data are reported as mean (standard deviation). The unipedal stance with eyes closed was very variable, both within and between subjects, so lower levels of reliability were observed for this test. Matched differences between the two sessions were close to zero on average, with individual performance varying.

	Static Bipedal			Static Unipedal
	Session 1	Session 2	Matched Difference	Session 1	Session 2	Matched Difference
AP Sway (cm)	0.41 (0.11)	0.46 (0.19)	0.02 (0.06)	1.73 (0.65)	1.76 (0.76)	0.02 (0.31)
ML Sway (cm)	0.18 (0.08)	0.17 (0.10)	−0.01 (0.04)	1.46 (0.55)	1.40 (0.58)	−0.02 (0.28)
Path Length (cm)	18.58 (7.02)	16.02 (6.29)	−1.01 (1.82)	169.08 (67.52)	150.95 (71.35)	−7.14 (13.12)
Max Path Velocity (cm/s)	2.64 (0.95)	2.48 (0.58)	−0.06 (0.49)	25.19 (22.27)	20.84 (16.17)	−1.71 (10.40)
95% Ellipse Area (cm2)	1.49 (0.95)	1.57 (1.52)	0.01 (0.21)	50.53 (36.14)	51.82 (43.14)	0.20 (6.34)
AP Entropy	0.47 (0.13)	0.40 (0.10)	−0.07 (0.11)	0.34 (0.17)	0.44 (0.18)	0.10 (0.23)
ML Entropy	0.28 (0.08)	0.24 (0.10)	−0.05 (0.08)	0.47 (0.11)	0.51 (0.11)	0.03 (0.11)
Renyi Entropy	3.65 (0.23)	3.58 (0.18)	−0.07 (0.33)	3.81 (0.17)	3.82 (0.16)	0.01 (0.15)
Shannon Entropy	2.61 (0.13)	2.57 (0.16)	−0.03 (0.10)	2.63 (0.14)	2.61 (0.07)	−0.03 (0.16)

DISCUSSION

Consistent with previous research, the BESS was observed to be associated with lower reliability than other assessments [26]. It was included in this balance protocol due to its ease of use and near ubiquity in the assessment of sports-related concussion. The learning effect that is often observed in studies involving the BESS was noted in this study as well, with subjects performing better in Session 2 than in Session 1 (Figure 1). The stances for the BESS are generally unique to study subjects, so additional administration of the BESS allows subjects to engage compensatory mechanisms to maintain postural control (Figure 2). Further, the one foot stance on a foam surface may be too difficult for subjects to complete, as all of the healthy, adult subjects in this study experienced 4 or more errors for this particular stance. Pediatric subjects with limited postural control would likely struggle to complete this task as well.

The SEBT has been commonly implemented to determine whether subjects are at risk for lower extremity injury. It is a reliable measure of dynamic postural control, which is a factor that may also be comprised during instances of concussion. One limitation associated with the SEBT is that it can be difficult for a single test administrator to determine whether a subject has shifted their balance towards their reaching foot while simultaneously seeing how far a subject has reached. Further, the overall SEBT can be a time-consuming assessment, with subjects having to complete three trials in eight different directions with both legs. A modified version of the SEBT, only involving reaches in the anterior, posterior lateral, and posterior medial directions, was employed in this study and was associated with high levels of reliability (Table 1). Using this modified SEBT with an active, pediatric population may be viable for assessing dynamic postural control, though further research is necessary. Assessing reliability and feasibility with a healthy youth population would be required, as well as utilizing the modified SEBT with concussed youth athletes to determine its ability to discriminate between healthy and concussed youth athletes.

Performance on the force plate varied between the four test conditions, though Renyi and Shannon Entropy were largely consistent across all tests (Tables 2 and 3). Matched differences between the two sessions were close to zero, which is representative of an assessment with little within-subject variation. Between-subject variability existed for the force plate metrics as represented by the standard deviations of computed metrics. The unipedal eyes closed test was very difficult for subjects to complete, with some having to place their second foot down at times in order to regain balance. Measures from this stance may not be representative of static postural control. The unipedal eyes open test required greater attention by subjects than the bipedal test in order to maintain postural control. This manifested itself in higher measures of all spatially varying metrics. With the frequency content of unipedal COP trajectories consisting of predominantly low frequency data (< 2 Hz), conducting this test with subject’s eyes open likely represents the best opportunity to differentiate postural control. The ability of a pediatric population to complete this test requires investigation, with the possibility of utilizing a shorter duration test for unipedal stances.

A good measure of balance would be one in which within-subject variability was low compared to between-subject variability. An individual subject’s score should not vary considerably from test to test in absence of some disturbance to postural control. Conversely, the measure should be sensitive enough to distinguish balance between subjects. The modified SEBT exhibited this favorable ratio of between-subjects to within-subjects variability, while the BESS did not (Figures 1 and 2). Individual performance varied considerably on the BESS between sessions, while performance on the modified SEBT was more stable between the two sessions. The force plate metrics used in this study were associated with high levels of between-subjects variability relative to within-subjects variability, specifically for the unipedal eyes open test.

Though the unipedal stances on the force plate and the modified SEBT employed in this study resulted in high levels of test-retest reliability, the tasks may not be feasible to apply to a pediatric population. The unipedal stance conducted with eyes closed was difficult for the healthy adult subjects in this study to complete, with a very unstable COP trajectory. It is likely that pediatric subjects who are still developing postural control would not be able to complete this static assessment. The anterior reaching stances of the modified SEBT resulted in lower reliability than either of the posterior reaching stances (Table 1). This task was also associated with the highest number of invalid trials, as subjects either transferred their weight to the reaching foot or could not maintain balance when attempting to return to their base of support. Use of this reaching stance with a pediatric population would potentially lead to lower levels of reliability and increase the total testing time. Previous research has discussed the feasibility challenges associated with using the SEBT with a pediatric group [27]. Using only those stances with the highest reliability from the modified SEBT would limit the overall testing time and subject fatigue, while providing the greatest opportunity for reliability with youth athletes.

CONCLUSIONS

A pilot study was conducted that utilized a series of clinical static and dynamic postural control assessments, in addition to unipedal and bipedal stances on an instrumented force plate. The BESS was associated with low levels of reliability and repeatability, whereas the modified SEBT tested consistently between the two sessions in this study. For the force plate testing, unipedal stances resulted in greater levels of variability between subjects. Conducting this test with the subject’s eyes closed may not be appropriate, or representative of a stable COP trajectory. This balance protocol will be further investigated in a pediatric population of youth football players. These assessments may be viable within this population, or youth-specific balance measures may be needed to appropriately assess the still developing postural control of youth athletes. Further, use of postural control testing in instances of pediatric concussion may better understanding of how this injury manifests in a younger population.