Sports-related concussion continues to garner attention in the media and among medical providers because an estimated 1.1 to 1.9 million young athletes sustain these injuries annually in the United States alone.1 Individuals who sustain a sports-related concussion exhibit an array of symptoms and deficits following injury, which often include impairments in postural control.2–5 Impaired postural control has been observed in 28.6% of athletes with concussion.6 It has been reported that sports-related concussion has a large effect on postural control up to 14 days after injury.7
Postural control assessments are commonly used for sports-related concussion assessment.8,9 Developments in technology in recent decades have led to a rise in instrumented strategies to evaluate postural control. These instrumented assessments provide quantifiable measures of postural control that may be more sensitive to identifying deficits following sports-related concussion compared to common clinical assessments with subjective scoring systems (eg, Balance Error Scoring System). Although clinically useful, especially in the acute phase following sports-related concussion, the subjective assessments may not be able to detect deficits that may linger or may not be obvious on clinical examination.
The Sensory Organization Test (SOT) is a commonly used assessment that removes or distorts specific sensory inputs to examine the sensory contributions to postural control.10 The SOT has identified deficits in postural control after sports-related concussion in collegiate athletes when compared to healthy matched controls.5 Although the SOT has been widely accepted and adopted by researchers and clinicians from several disciplines,11 further improvements in technology have implemented immersive virtual reality components to the SOT. The immersive virtual reality environment uses the same center of pressure and sway measurements to quantify postural control, but replaces the traditional mechanical surround system with an immersive virtual reality dome. The immersive virtual reality system uses an endless tunnel virtual environment instead of a mobile mechanical surround. The virtual reality surround creates the illusion of self-motion through the full-field immersion without reference points or auditory cueing created by a mechanical surround.12,13 Data collected using the immersive virtual reality surround have demonstrated similar norms to those collected on the previous system with the mechanical surround.12 However, despite comparisons of normative data in both systems by True-blood et al.,12 it is unknown how this change may affect the reliability characteristics of the assessment.
A previous study in older adults determined that three trials should be collected for each SOT condition to achieve reliable test–retest results.14 Further examining the number of recommended trials for each SOT condition may reduce the number of required trials and better account for changes in test trials as a result of familiarization. If only two trials were needed for reliable results, testing time could be reduced by at least 4 minutes per patient, which may be particularly beneficial when collecting preinjury baseline measurements on a large number of athletes. Calculating additional statistics such as the minimal detectable change (MDC) and mean change scores across testing sessions may help provide information regarding the use of the test and interpretation of the results. Therefore, this information could be beneficial for clinical and research settings by reducing testing time and enhancing the ability to interpret the results.
The test–retest reliability and MDC values have not been established. Establishing these properties is important to ensure consistent data can be obtained across testing sessions, which is critical for making accurate conclusions from the data. Therefore, the purpose of this study was to determine the test–retest reliability and MDC of the SOT in healthy young adults using the BBA-CDP. A secondary purpose was to determine the optimum number of trials for the most reliable performance. It was hypothesized that the SOT administered using the BBA-CDP would exhibit good to excellent reliability across testing sessions and reasonable MDC values in all SOT conditions.
A convenience sample of men and women aged 18 to 35 years from the general population of a large public university were recruited to participate over a 6-month period. A total of 41 healthy adults were recruited, screened for eligibility, and enrolled in the study. Thirty-nine participants completed both sessions of testing (22 female, 17 male; mean age: 26.35 ± 4.28 years) and were included in the final sample. Participants were considered healthy individuals with no history of concussion or lower extremity injury in the previous year, per self-report. Participants were not currently taking any medications or experiencing any known health conditions (eg, peripheral neuropathy, motion sickness, vertigo, or epilepsy) that could influence postural control. All participants were given the opportunity to ask questions or gain clarity before signing a written informed consent form that was approved by the university's institutional review board.
The SOT was administered using the Bertec Balance Advantage-CDP (BBA-CDP; Bertec Corp., Columbus, OH) (Figure 1) equipped with Bertec Balance Advantage software, which was used in the collection and data reduction of the SOT. The BBA-CDP uses 18” × 20” dual force plates to measure sway, sway velocity, and weight shift. The dynamic force plate system is encircled by a dome-shaped surround with a projector to create the impression of a moving immersive virtual reality environment. The dynamic quality of the force plate rotated exclusively in the anterior and posterior plane, matching that of the participant. Similarly, the virtual environment responded to the sway of the participant by providing the impression of the light pattern moving toward or away from the participant, challenging the systems at work. The SOT protocol with six conditions (Table 1) and three 20-second trials of each condition was used on the BBA-CDP.
BBA-CPD device (Bertec Corp., Columbus, OH).
Sensory Organization Test Conditions
After providing written informed consent, participants completed a demographics questionnaire that included age, height, mass, limb dominance, and additional questions to ensure they met the criteria for inclusion. To begin the SOT, participants were secured in a safety harness and provided information regarding foot placement and procedures before stepping into the BBA-CDP. The lights in the laboratory were dimmed to a standardized setting using a digital Leaton Luxmeter L830b (Shenzhen DeXi Electronics Co., Ltd, Shenzhen, China) to 45 ± 5 lx, to ensure consistent lighting to enhance the visual field projected for the SOT. Participants were given standardized instructions before beginning each SOT condition and proceeded through testing from Condition 1 through Condition 6, subsequently. All tests were administered based on the manufacturer's recommendations in a quiet, distraction-free laboratory setting. Following the initial testing session, participants returned 1 week later (7 ± 1 days) to complete the second testing session. During the second testing session, the SOT was administered using the same procedure as described for the first testing session by the same researcher.
During each testing session, participants completed the six conditions of the SOT in the same order of the original standardized SOT. Each of the six conditions were tested through three 20-second trials. The SOT manipulated the vestibular and somatosensory systems through the immersive virtual environment and the sway referenced feedback from the moving force plates, respectively.
For each trial of the SOT, an equilibrium score was calculated. The equilibrium score is the angular difference between the participant's calculated maximum anterior-posterior center of gravity displacements and the theoretical range of normal sway, or 12.5°.15 Higher equilibrium scores indicated better postural control based on fewer degrees of sway. Lower equilibrium scores represented poorer postural control based on greater degrees of sway, and the score is presented as a percentage.
Descriptive statistics were calculated for the equilibrium scores for each SOT condition along with the mean difference across testing sessions. Intraclass correlation coefficients (ICC 2, 1) with corresponding 95% confidence interval (CI) were used to measure the reliability of the scores for each condition based on the first trial only, the average of the first two trials, the average of the last two trials, and the average of all three trials. ICCs were interpreted based on the guidelines reported in Koo and Li16 as poor (< 0.49), moderate (0.50 to 0.74), good (0.75 to 0.89), or excellent (> 0.90). Additionally, the standard error of measurement (SEM) and MDC were calculated to determine the level of change required in equilibrium scores to surpass measurement error. The MDC95 was calculated by multiplying the SEM by the square root of two by the z-score of 1.96 to reflect a confidence interval of 95%17: MDC = SEM × √2 × 1.96.
In addition to reliability statistics, paired t tests were used to compare each SOT condition across testing sessions. Corresponding standardized response mean (SRM) effect sizes were calculated by dividing the mean change scores from each participant by the standard deviation of the change scores:
Finally, the stability of the individual trials for each SOT condition was assessed using separate repeated-measures analyses of variance with post-hoc paired tests in the event of a significant main effect. This analysis was completed for each SOT condition in the first and second testing sessions. For all analyses, the level of significance was set at a P value of less than .05.
All statistical analyses were performed using SPSS software (version 24.0; IBM Corporation, Armonk, NY).
A total of 39 participants completed the study. Two participants were lost to follow-up. The descriptive statistics, ICC, SEM, and MDC for each SOT condition are displayed in Table 2 for the first trial only, the mean of the first two trials, the mean of the last two trials, and the mean of all three trials.
SOT Equilibrium Score Means of the Trials
The test–retest reliability of the first trial for each SOT condition ranged from poor to good (ICC = 0.44 to 0.77). However, the reliability of the mean of the first two trials for each condition demonstrated moderate to good reliability (ICC = 0.69 to 0.89). Similarly, the mean of the last two trials for each condition exhibited moderate to good reliability (ICC = 0.72 to 0.88). The strongest test–retest reliability was achieved by using the mean of all three trials, which resulted in good to excellent reliability (ICC = 0.81 to 0.91).
When comparing SOT performance across testing sessions, significant improvements with moderate effect sizes were identified in the second testing session for Conditions 4, 5, and 6 (P ≤ .05); regardless of the combination of trials (Table 2). When examining differences across trials within each SOT condition, the first trial was significantly lower than either the second or third trial for Conditions 2, 3, 4, and 6 during the first testing session (Table 3). However, only Condition 3 demonstrated a difference between the first and third trial during the second data collection session (Table 3).
Individual Trial Descriptive Statistics and Omnibus Outcomes for RM ANOVA
The purpose of this study was to determine the test–retest reliability and MDC of the SOT in healthy young adults using the BBA-CDP. The primary finding of this study was that all six SOT conditions exhibited satisfactory reliability across two testing sessions 1 week apart and examined with at least two trials. Further analysis revealed that the average of three trials had the greatest reliability among all combinations tested in this study. Consequently, the average of all three trials also resulted in lower MDC values compared to all other trial combinations. However, systematic improvements were identified in SOT performance across testing sessions, predominantly in the more challenging SOT conditions, and performance across trials varied during the first testing session. These results help outline the application of the SOT in a young healthy population through better understanding of changes across repeated administrations.
The current study examined the reliability of each SOT condition through analysis of the first trial and various trial combinations. It was hypothesized that the SOT administered using the BBA-CDP would exhibit good to excellent reliability when a minimum of two trials were used for analysis. The best reliability occurred with the mean of all three trials for each condition, as recommended by the manufacturer. However, the reliability for the mean of the first two trials or the last two trials was generally sufficient and only marginally different than the mean of three trials. Using only the first trial for each condition did not demonstrate sufficient reliability across all conditions and is therefore not recommended for clinical or research settings.
These findings are supported by a previous study14 that identified insufficient reliability with the first trial and adequate reliability with the mean of all three trials across the SOT conditions using a different system. However, the reliability demonstrated in the current study was stronger than previously reported. This may be attributed to the inclusion of a healthy young adult sample or instrumentation differences associated with the BBA-CDP, such as the virtual reality dome, which no longer exhibited the motorized sound associated with mechanical surrounds.
It is important to note that the equilibrium score means and standard deviations reported in this investigation closely match those reported in a previous investigation using CPD with immersive virtual reality in age groups represented in this investigation.12 Additionally, similar trends across the condition equilibrium scores were noted. These similarities confirm our confidence in the clinical applicability and consistency of the results presented.
Previous reliability studies, with similar technologies without the virtual environment, yielded lower reliability statistics (0.26 to 0.64).10,14 One of the aforementioned investigations of reliability used older adults. This may have an effect on the ICC values because age and neuromuscular changes may play a role in an individual's ability to perform the assessment.14 The investigation in healthy young adults yielded a higher ICC (0.64) for the composite score that would include all trials and conditions.10 Yet, the differences in the virtual environment must still be considered. Overall, the BBA-CDP demonstrated moderate to excellent test–retest reliability in healthy young individuals when a minimum of at least two trials are recorded for each SOT condition.
Among the most notable effects was the systematic increase in SOT performance observed between testing sessions 1 and 2, particularly for Conditions 4 to 6. The less challenging conditions (Conditions 1 to 3) demonstrated relatively small changes across sessions compared to the more challenging conditions (Conditions 4 to 6), which demonstrated larger improvements in performance in session 2. For example, 79.5% of participants demonstrated improved equilibrium scores for Condition 6 when using the average of all three trials. These findings suggest that participants may have been more familiar or able to cope more effectively with the manipulation of multiple sensory conditions on the re-test. This systematic improvement in performance was evident in each series of analyses, suggesting a potential bias due to learning or other potential adaptations, which may have occurred for the more difficult conditions. Similarly, in an evaluation of the learning effect of repeated administrations of the SOT in a similar sample, Conditions 4, 5, and 6 saw improvements over three administrations that plateaued for the final two administrations, suggesting the need for multiple baselines to achieve the most stable and accurate measure.10
Additionally, the participants generally achieved high equilibrium scores during the first three conditions, which may have left less room for improvement on the retest session compared to the more challenging conditions that had lower starting equilibrium scores and a greater opportunity for improvement in the second testing session. Previous research10 has suggested the need for repeated/familiarization baseline testing due to the changes in the mean over repeated administrations, as observed in our study. Previously, the reliability of the SOT increased with a fourth, fifth, and sixth testing session. Therefore, certain SOT conditions have the potential to exhibit improvements in performance from repeated assessments when captured using the BBA-CDP. Future studies should examine whether this effect can be mitigated through additional practice trials or repeated baselines to enhance familiarization with the test.
An additional trend observed from this analysis was within the MDC values. Across each of the analyses, the MDC values tended to be the greatest in Conditions 4 to 6, indicating larger magnitudes of change are needed to exceed the measurement error, learning effect, or other adaptations that may be occurring across repeated administrations for this condition. Conditions 4 to 6 may be most susceptible to improvements over time because they involve the moving platform or challenges to multiple sensory systems, increasing the level of constraint on the participant. Generally, MDC values decreased across most conditions with the addition of more trials. This trend in MDC values was likely related to the improvements in reliability associated with increasing the number of trials. When applying these MDC values in clinical or research settings, it is important to note that relatively small changes in equilibrium scores may be relevant in Conditions 1 to 3 (3.01 to 5.14), whereas greater changes will be necessary in Conditions 4 to 6 (8.88 to 30.81) to identify changes in individual patients that surpass measurement error. These values should be reevaluated in clinical populations but may serve as guidance for interpreting changes in SOT scores in a young healthy adult population.
Finally, the trial-to-trial stability for several of the SOT conditions requires further consideration. During the first testing session, the first trial of SOT Conditions 2, 3, 4, and 6 was significantly lower than either the second or third trial. However, no differences were identified between the second or third trial. Additionally, only Condition 3 exhibited differences in the first trial during the second test session. These findings suggest that during a participant's first time completing the SOT, the first trial for most conditions may not provide an accurate representation of postural control. This may be easily remedied by adding a practice trial to the testing protocol to increase familiarity with the testing conditions.
This study was not without limitations. First, the study was performed with a healthy young adult sample. This sample was selected because we anticipated that performance would remain stable in this group. However, although self-reported as healthy individuals, the physical activity status of the population was unknown. The results of these analyses may not be representative of patient populations with postural control impairments due to clinical conditions, across different age groups, or across different levels of athletic ability. To better understand the utility of the SOT in individuals with sports-related concussion, it is important to evaluate the assessment in that unique population. In addition, we observed a potential learning effect or adaptation with repeated administration of the SOT when using the BBA-CDP; however, we were unable to determine how many testing sessions would be required for performance to stabilize. It is recommended that multiple baselines or added practice trials be considered to establish a representative baseline performance, particularly for Conditions 4 to 6. Repeating baselines 1 week after the original baseline, as done in this study, and treating the original as a practice trial may improve performance on this assessment and provide a better measure of the true baseline. Finally, participants were secured in a harness for all testing, as recommended by the manufacturer. It is possible that the harness affected the overall stability of the individual by limiting the amount of natural motion or providing a stabilizing force, particularly in conditions that used the dynamic surround or moving platform.
Implications for Clinical Practice
The main finding of this study was that the greatest reliability across all SOT conditions was achieved when all three test trials were included for repeated testing. Additionally, improved performance should be expected for several SOT conditions with repeated testing and MDC values should be applied to identify changes beyond measurement error or time effects. Specifically, these results indicate that additional familiarization, particularly with Conditions 4 to 6, captures data representative of the patient or participant during initial assessment. Practice trials are also recommended for a majority of the SOT conditions, particularly when it is the individual's first SOT. Additionally, the results of this study using the BBA-CPD SOT proved more reliable than the previous literature using the NeuroCom SmartEqui Test, suggesting it may serve as a better clinical tool.10 The results of this research may assist clinicians in understanding SOT scores and values in a similar population and the reliability of those scores with a repeated administration. It may also suggest the best practices when baseline testing a young healthy adult population and assist with interpreting SOT performance following injury.
The results of this study support the use of the BBACDP SOT protocols as a reliable measure of postural control in a healthy young adult sample. In addition, the results of this study provide methodological considerations to improve clinical application and interpretation for clinicians when assessing postural control using the SOT. The results support the utility of the SOT as a tool for objectively measuring baseline postural control in athletic populations at risk of sports-related concussion to better understand potential deficits should an injury occur. Future studies should evaluate the reliability of the SOT in pathological populations and strategies to mitigate potential learning effects.
- Bryan MA, Rowhani-Rahbar A, Comstock RD, Rivara FSeattle Sports Concussion Research Collaborative. Sports- and recreation-related concussions in US youth. Pediatrics. 2016;138(1):e20154635. doi:10.1542/peds.2015-4635 [CrossRef]
- Guskiewicz KM. Postural stability assessment following concussion: one piece of the puzzle. Clin J Sport Med. 2001;11(3):182–189. doi:10.1097/00042752-200107000-00009 [CrossRef]
- Guskiewicz KM, Perrin DH, Gansneder BM. Effect of mild head injury on postural stability in athletes. J Athl Train. 1996;31(4):300–306.
- Riemann BL, Guskiewicz KM. Effects of mild head injury on postural stability as measured through clinical balance testing. J Athl Train. 2000;35(1):19–25.
- Guskiewicz KM, Ross SE, Marshall SW. Postural stability and neuropsychological deficits after concussion in collegiate athletes. J Athl Train. 2001;36(3):263–273.
- Valovich McLeod TC, Hale TD. Vestibular and balance issues following sport-related concussion. Brain Inj. 2015;29(2):175–184. doi:10.3109/02699052.2014.965206 [CrossRef]
- Broglio SP, Puetz TW. The effect of sport concussion on neurocognitive function, self-report symptoms and postural control: a meta-analysis. Sports Med. 2008;38(1):53–67. doi:10.2165/00007256-200838010-00005 [CrossRef]
- Guskiewicz KM, Bruce SL, Cantu RC, et al. National Athletic Trainers' Association Position Statement: management of sport-related concussion. J Athl Train. 2004;39(3):280–297. doi:10.1227/01.NEU.0000143800.49798.19 [CrossRef]
- McCrory P, Meeuwisse W, Dvorak J, et al. Consensus statement on concussion in sport—the 5th international conference on concussion in sport held in Berlin, October 2016. Br J Sports Med. 2017:bjs-ports-2017-097699.
- Wrisley DM, Stephens MJ, Mosley S, Wojnowski A, Duffy J, Burkard R. Learning effects of repetitive administrations of the sensory organization test in healthy young adults. Arch Phys Med Rehabil. 2007;88(8):1049–1054. doi:10.1016/j.apmr.2007.05.003 [CrossRef]
- Guskiewicz KM, Perrin DH. Research and clinical applications of assessing balance. J Sport Rehabil. 1996;5(1):45–63. doi:10.1123/jsr.5.1.45 [CrossRef]
- Trueblood PR, Rivera M, Lopez C, Bentley C, Wubenhorst N. Age-based normative data for a computerized dynamic posturography system that uses a virtual visual surround environment. Acta Otolaryngol. 2018;138(7):597–602. doi:10.1080/00016489.2018.1429653 [CrossRef]
- Akizuki H, Uno A, Arai K, et al. Effects of immersion in virtual reality on postural control. Neurosci Lett. 2005;379(1):23–26. doi:10.1016/j.neulet.2004.12.041 [CrossRef]
- Ford-Smith CD, Wyman JF, Elswick RK Jr, Fernandez T, Newton RA. Test–retest reliability of the sensory organization test in noninstitutionalized older adults. Arch Phys Med Rehabil. 1995;76(1):77–81. doi:10.1016/S0003-9993(95)80047-6 [CrossRef]
- Cohen H, Heaton LG, Congdon SL, Jenkins HA. Changes in sensory organization test scores with age. Age Ageing. 1996;25(1):39–44. doi:10.1093/ageing/25.1.39 [CrossRef]
- Koo TK, Li MY. A guideline of selecting and reporting intra-class correlation coefficients for reliability research. J Chiropr Med. 2016;15(2):155–163. doi:10.1016/j.jcm.2016.02.012 [CrossRef]
- Haley SM, Fragala-Pinkham MA. Interpreting change scores of tests and measures used in physical therapy. Phys Ther. 2006;86(5):735–743.
Sensory Organization Test Conditions
SOT Equilibrium Score Means of the Trials
|SOT Condition||Time 1 (Mean ±SD)||Time 2 (Mean ±SD)||ICC (2,1)||SEM||MDC95||MD||P||SRM|
| 1||92.41 ±4.47||92.08 ±5.95||0.70||2.88||7.98||−0.33 ±4.16||.620||−0.08|
| 2||91.05 ± 4.74||90.59 ± 12.80||0.44||6.55||18.14||−0.46 ±10.32||.781||−0.04|
| 3||89.64 ± 4.50||90.38 ± 4.92||0.63||2.85||7.89||0.74 ±4.08||.262||0.18|
| 4||78.05 ± 10.91||81.49 ± 8.21||0.45||7.06||19.58||3.44 ±10.22||.042||0.34|
| 5||65.05 ± 14.13||72.05 ± 14.03||0.77||6.71||18.60||7.00 ±12.27||.001||0.57|
| 6||61.54 ± 16.77||68.72 ± 17.32||0.58||11.11||30.80||7.18 ±15.92||.008||0.45|
|Trials 1 and 2|
| 1||92.59 ±3.85||92.35 ±4.58||0.89||1.42||3.93||−0.24 ±2.73||.581||−0.09|
| 2||91.85 ±3.37||91.79 ±6.98||0.71||2.78||7.70||−0.05 ±5.25||.952||−0.01|
| 3||90.40 ±3.56||90.97 ±4.44||0.73||2.09||5.80||0.58 ±3.78||.347||0.15|
| 4||79.58 ±7.95||82.21 ±7.36||0.69||4.27||11.83||2.63 ±7.56||.036||0.35|
| 5||65.62 ±13.88||71.55 ±12.22||0.81||5.69||15.77||5.94 ±10.58||.001||0.56|
| 6||65.53 ±13.55||71.28 ±13.82||0.78||6.46||17.90||5.76 ±11.85||.004||0.49|
|Trials 2 and 3|
| 1||93.14 ±3.03||92.74 ±3.33||0.88||1.10||3.04||−0.40 ±2.10||.245||−0.19|
| 2||92.29 ±2.92||92.73 ±2.77||0.76||1.39||3.86||0.44 ±2.59||.321||0.17|
| 3||91.51 ±3.93||91.83 ± 4.26||0.75||2.09||5.70||0.32 ±3.73||.595||0.09|
| 4||81.28 ±7.96||82.82 ±6.63||0.81||3.20||8.88||1.54 ±5.97||.250||0.26|
| 5||66.29 ±16.01||72.41 ±9.98||0.72||6.85||18.99||6.12 ±12.61||.004||0.49|
| 6||68.65 ±13.21||72.53 ±12.81||0.83||5.32||14.74||3.87 ±9.97||.020||0.39|
|Trials 1, 2, and 3|
| 1||92.90 ±3.26||92.52 ±3.94||0.91||1.09||3.01||−0.38 ±2.11||.282||−0.18|
| 2||91.88 ±3.25||92.02 ±5.65||0.83||1.85||5.13||0.14 ±3.58||.813||0.04|
| 3||90.89 ±3.61||91.35 ±4.15||0.84||1.54||4.26||0.46 ±2.09||.327||0.22|
| 4||80.21 ±8.24||82.38 ±6.79||0.81||3.24||8.99||2.17 ±6.06||.031||0.36|
| 5||65.88 ±13.95||72.29 ± 10.68||0.84||4.93||13.66||6.41 ±9.35||.001||0.69|
| 6||66.28 ±13.38||71.26 ±12.98||0.86||4.95||13.70||4.97 ±9.38||.002||0.53|
Individual Trial Descriptive Statistics and Omnibus Outcomes for RM ANOVA
|Condition||Trial 1 (Mean ± SD)||Trial 2 (Mean ± SD)||Trial 3 (Mean ± SD)||F (df)||P||Partial Eta Squared|
| 1||92.41 ± 4.47||92.77 ± 3.96||93.51 ± 2.54||2.47 (2, 76)||.091||0.06|
| 2||91.05 ± 4.81||92.64 ± 3.04a||91.95 ± 3.82||3.43 (2, 76)||.038||0.08|
| 3||89.64 ± 4.56||91.15 ± 4.39||91.87 ± 4.72a||4.58 (2, 76)||.013||0.11|
| 4||78.05 ± 11.05||81.10 ± 7.96||81.46 ± 10.46a||3.19 (2, 76)||.047||0.08|
| 5||65.05 ± 14.31||66.18 ± 18.78||66.41 ± 16.71||0.17 (2, 76)||.841||0.01|
| 6||61.54 ± 16.99||69.51 ± 13.13a||67.79 ± 16.31a||7.81 (2, 76)||.001||0.17|
| 1||92.08 ± 6.03||92.62 ± 3.93||92.87 ± 3.43||0.81 (2, 76)||.447||0.02|
| 2||90.59 ± 12.96||93.00 ± 2.70||92.46 ± 3.60||1.39 (2, 76)||.256||0.04|
| 3||90.38 ± 4.98||91.56 ± 4.94||92.10 ± 4.52a||3.63 (2, 76)||.031||0.09|
| 4||81.49 ± 8.32||82.92 ± 7.74||82.72 ± 7.01||1.30 (2, 76)||.278||0.03|
| 5||72.05 ± 14.22||71.05 ± 12.63||73.77 ± 9.59||1.44 (2, 76)||.244||0.04|
| 6||68.72 ± 17.54||73.85 ± 14.00||71.21 ± 14.81||2.52 (2, 76)||.088||0.06|