- Email: [email protected]
Construct validity of the BESTest, mini-BESTest and briefBESTest in adults aged 50 years and older
G Model
GAIPOS-4507; No. of Pages 5 Gait & Posture xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Construct validity of the BESTest, mini-BESTest and briefBESTest in adults aged 50 years and older Sachi O’Hoski a, Kathryn M. Sibley b,c, Dina Brooks a,b,c, Marla K. Beauchamp a,d,* a
West Park Healthcare Centre, Respiratory Medicine, 82 Buttonwood Avenue, Toronto, ON, M6M 2J5, Canada University of Toronto, Department of Physical Therapy, Faculty of Medicine, 500 University Avenue, Toronto, ON M5G 1V7, Canada c Toronto Rehabilitation Institute – University Health Network, 550 University Avenue, Toronto, ON M5G 2A2, Canada d Department of Physical Medicine and Rehabilitation, Harvard Medical School, Spaulding Rehabilitation Hospital, Cambridge, MA, USA b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 22 November 2014 Received in revised form 5 June 2015 Accepted 15 June 2015
Background: The Balance Evaluation Systems Test (BESTest) and its two abbreviated versions (miniBESTest and briefBESTest) are functional balance tools that have yet to be validated in middle aged and elderly people living in the community. Objective: Determine the construct validity of the three BESTest versions by comparing them with commonly-used measures of balance, balance confidence and physical activity, and examining their ability to discriminate between groups with respect to falls and fall risk. Methods: This was a secondary analysis of data from 79 adults (mean age 68.7 10.57 years). Pearson correlation coefficients were used to examine the relationships between each BESTest measure and the Activities-Specific Balance Confidence (ABC) scale, the Physical Activity Scale for the Elderly (PASE), the Timed Up and Go (TUG) and the Single Leg Stance (SLS) test. Independent t-tests were used to examine differences in balance between fallers (1 fall in previous year) and non-fallers and individuals classified at low versus high fall risk using the Elderly Falls Screening Test (EFST). Results: The BESTest measures showed moderate associations with the ABC scale and TUG (r = 0.62–0.67 and 0.60 to 0.68 respectively), fair associations (r = 0.33–0.40) with the PASE and moderate to high associations (r = 0.67–0.77) with the SLS. Fallers showed a trend (p = 0.054) for lower scores on the original BESTest, and people at high risk for falls had significantly lower scores on all BESTest versions. Conclusions: These findings support the construct validity of the BESTest, mini-BESTest and briefBESTest in adults over 50 years old. ß 2015 Published by Elsevier B.V.
Keywords: Postural balance Geriatric assessment Outcomes assessment Aged
Falls in later adulthood are a significant public health problem [1]. While numerous fall risk factors have been identified, balance impairment is recognized as one of the most important modifiable risk factors for falls [2]. The ability to maintain balance is a complex skill that requires the integration of information from multiple physiological systems [3], many of which are impaired with age [2]. Clinical measures of balance with sound psychometric properties are needed to screen for balance problems in order to guide treatment and inform fall prevention strategies. The Balance Evaluation Systems Test (BESTest) is a clinical balance tool that targets six subsystems of postural control (see Box 1) in order to identify the underlying impairments contributing
* Corresponding author at: Department of Physical Medicine and Rehabilitation, Harvard Medical School, Spaulding Outpatient Center Cambridge, 1575 Cambridge St., Cambridge, MA 02138, USA. Tel.: +1 617 952 6954; fax: +1 617 952 6965. E-mail addresses: [email protected], [email protected] (M.K. Beauchamp).
to dysfunctional balance [4]. In a recent review, it was identified as the only standardized balance measure that evaluates all components of balance consistent with established conceptual models [5]. The BESTest has been used in a variety of populations [4,6] and has been shown to have strong psychometric properties in people with Parkinson’s disease [7]. However, few studies have reported on its use in a general population of middle-aged and older adults [4,6,8]. Two abbreviated versions of the BESTest (mini-BESTest and briefBESTest) have been developed as alternatives to the original test. These abbreviated versions take 10–20 min to complete [9,10], considerably less than the 30–60 min to administer the BESTest [4,9,10], time that is not feasible in many clinical settings. The miniBESTest was developed using factor analysis to identify the items of the BESTest that represented dynamic balance. Rasch analysis was then used to improve the rating categories and eliminate some items [10]. The result was a 14 item test of dynamic balance [10]. The miniBESTest has been used in several clinical populations [10] and evidence for its psychometric properties has been shown in people
http://dx.doi.org/10.1016/j.gaitpost.2015.06.006 0966-6362/ß 2015 Published by Elsevier B.V.
Please cite this article in press as: O’Hoski S, et al. Construct validity of the BESTest, mini-BESTest and briefBESTest in adults aged 50 years and older. Gait Posture (2015), http://dx.doi.org/10.1016/j.gaitpost.2015.06.006
G Model
GAIPOS-4507; No. of Pages 5 S. O’Hoski et al. / Gait & Posture xxx (2015) xxx–xxx
2
Box 1. Description of categories and items of the Balance Evaluation Systems Test. Category
Description and examples of tasks
Biomechanical constraints
Evaluates constraints on standing balance including postural alignment, ankle range of motion and hip strength Evaluates ability to move the body over its base of support by leaning forward and laterally and evaluates ability to return to gravitational vertical Evaluates active movement of the center of mass in anticipation of performing sit to stand, Single Leg Stance, and stair tap Evaluates in-place and compensatory stepping responses to external perturbations from the front, side and back Evaluates increases in postural sway under different sensory conditions such as standing on flat ground or foam with eyes open or closed Evaluates stability while walking under different conditions such as changing speed, looking from side to side and stepping over an obstacle
Stability limits/verticality
Anticipatory postural adjustments Postural responses
Sensory orientation
Stability in gait
with Parkinson’s disease [7]. Unlike the mini-BESTest, the briefBESTest maintains the original theoretical basis of the BESTest; it was created using item-total correlations to identify the most representative item from each subsection of the BESTest [9]. The brief-BESTest has similar inter-rater reliability to the BESTest and mini-BESTest and is better able to discriminate between fallers and non-fallers with and without a neurological diagnosis [9]. The construct validity of the BESTest, mini-BESTest and briefBESTest has not yet been examined in a general community-based sample. Therefore, our primary objective was to determine the associations between the BESTest and its abbreviated versions and measures of other theoretically related constructs (i.e., convergent validity) including measures of balance, balance confidence and physical activity in adults aged over 50 years living in the community. We hypothesized that all three BESTest scores would have high correlations with self-reported balance confidence and other performance-based measures of balance and moderate correlations with physical activity. In addition, while the BESTest, mini-BESTest and briefBESTest have been shown to discriminate between fallers and non-fallers with Parkinson’s disease [7,11], this has not been examined in those without neurological conditions. Therefore, our secondary objective was to determine the ability of the BESTest, mini-BESTest and briefBESTest to discriminate between groups with respect to fall history and risk of falls and to compare the discriminative ability of the BESTest measures with that of commonly used balance measures [12]; the self-reported Activities-Specific Balance Confidence (ABC) scale [13], and the Timed Upand-Go (TUG) test [14] and Single Leg Stance (SLS) test [14]. We hypothesized that fallers and those classified as being at high risk of falls would score significantly lower on all balance measures. 1. Methods This was a secondary analysis of a previous study; the methods have been described in detail elsewhere [8]. Briefly, the study protocol was approved by the University Research Ethics Board and written informed consent was obtained at the beginning of each data collection session. Individuals who met the following criteria were recruited: (1) age between 50 and 89 years, (2) residing independently in the community, (3) able to understand and follow three-step instructions, (4) able to walk 6 m independently
(without a gait aid). Participants were excluded if they reported: (1) a history of dizziness or fainting, (2) the presence of any cardiorespiratory, neurological or musculoskeletal condition that severely affected their balance (e.g., chronic obstructive pulmonary disease, stroke, recent hip or knee replacement), or (3) the use of medication(s) that they felt caused dizziness or affected their balance (e.g., antidepressants). Data were collected between January and July 2012 by raters trained in the administration of the BESTest. Demographic data (sex, age, height, weight) and all written questionnaires were collected before administering the BESTest. Two testers were present for each testing session; one to observe and score the items and one to supervise for safety. Scoring of the abbreviated tests was completed after the data collection session based on performance of the original BESTest tasks. 1.1. Outcome measures 1.1.1. Balance Evaluation Systems Test (BESTest) The BESTest is a 36 item test comprised of six subsections that contribute to postural control (see Box 1) [4]. Each item is scored from 0 to 3 points based on time or performance criteria resulting in a total possible score of 108 points, which is converted to a percentage score. Higher scores indicate better balance. 1.1.2. mini-BESTest The mini-BESTest is a 14-item test of dynamic balance that includes tasks from the BESTest subsystems ‘anticipatory postural adjustments’, ‘postural responses’, ‘sensory orientation’ and ‘stability in gait’ [10]. Each item is scored from 0 to 2 points resulting in a total score out of 28 points with higher scores indicating better balance. 1.1.3. briefBESTest The briefBESTest is an eight-item test comprised of one item from each subsection of the BESTest, with two items (SLS and functional reach forward) scored bilaterally [9]. Each item is scored from 0 to 3 points with a total possible score out of 24 points (higher scores for better balance). 1.1.4. Activities-Specific Balance Confidence (ABC) scale The ABC scale is a 16-item self-report questionnaire requiring individuals to indicate their confidence in completing progressively difficult tasks without becoming unsteady or losing their balance (0% = no confidence; 100% = completely confident) [13]. The percentage assigned for each task is summed and divided by 16 in order to obtain the overall score. The ABC scale has good test–retest reliability, convergent and criterion validity [13] and the ability to discriminate between fallers and non-fallers in community-dwelling elderly (cut-off score 67%) [15]. 1.1.5. Physical Activity Scale for the Elderly (PASE) The PASE consists of 26 questions that assess an individual’s physical activity level over the previous 7 days [16]. Subscale categories include leisure time, household and occupational physical activity. The frequency (never, seldom, sometimes, often) and duration (hours) of physical activity is recorded. Responses to individual items are weighted and summed to calculate the subscores and total score, which can range from 0 to 400 points. A high score is indicative of a high level of physical activity. The PASE has good test–retest reliability [16] and good construct validity when scores are compared to measures of static balance [16] and accelerometer data [17]. 1.1.6. Timed Up and Go (TUG) test Participants are instructed to stand up from a chair, walk 3 m at their usual walking speed, turn 1808, walk back to the chair and sit
Please cite this article in press as: O’Hoski S, et al. Construct validity of the BESTest, mini-BESTest and briefBESTest in adults aged 50 years and older. Gait Posture (2015), http://dx.doi.org/10.1016/j.gaitpost.2015.06.006
G Model
GAIPOS-4507; No. of Pages 5 S. O’Hoski et al. / Gait & Posture xxx (2015) xxx–xxx
down [18]. The entire task is timed and a lower time to completion indicates better mobility [18]. The TUG has been shown to have excellent intra- and inter-rater reliability and good convergent validity with other measures of balance [14]. While the traditional cut-off (13.5 s) has good specificity for ‘‘ruling in’’ falls, a recent metaanalysis noted it had low sensitivity and limited ability to discriminate between prospective fallers and non-fallers in community-dwelling older adults (area under the curve (AUC) = 0.57) [19]. 1.1.7. Single Leg Stance (SLS) test For this task, participants are asked to place their hands on their hips, lift one foot off the floor and hold that position for as long as possible. Higher times indicate better balance performance. The best time on each leg was recorded but only the highest time achieved was used for data analysis. SLS time has been shown to have excellent test–retest reliability [14] and to be able to discriminate between fallers and non-fallers in community dwelling elderly (AUC = 0.64) [14]. 1.1.8. Elderly Falls Screening Test (EFST) The EFST is a five-item test that combines self-reported fall questions with a gait evaluation (quality and speed) to determine fall risk [20]. Respondents are asked to indicate whether or not they have experienced a fall in the last year, defined as finding themselves ‘‘suddenly on the ground, without intending to get there, after [they] were in either a lying, sitting or standing position’’. For this study we classified ‘fallers’ as those who responded ‘‘yes’’ and ‘non-fallers’ as those who responded ‘‘no’’. The EFST total score ranges between 0 (low fall risk) and 5 (high fall risk), with participants receiving one point for each of the following: (1) two or more falls in the past year, (2) any injury from a fall, (3) occasional or frequent near falls, (4) taking longer than 10 s to walk 5 m, and (5) uneven, shuffling, wide-based, or unsteady gait. Those with total score of 2 points are considered at high-risk of future falls [20]. The EFST has good criterion and predictive validity for falls in the elderly [20]. 1.2. Data analysis Descriptive statistics (mean, SD, range) were calculated for age, height, weight, body mass index and each of the measures. Pearson correlation coefficients were calculated to examine the relationship between the BESTest, mini-BESTest, and briefBESTest, and the ABC, PASE, TUG, and SLS. By convention, a correlation of 0.00–0.25 was interpreted to indicate little or no relationship, 0.25–0.5 as a fair relationship, 0.5–0.75 as moderate, and above 0.75 as a very good to excellent relationship. Independent t-tests were used to examine differences between fallers versus non-fallers and individuals at low versus high fall risk based on the EFST. All analyses were conducted with SPSS (version 19.0 for Windows; SPSS Inc.; Chicago, United States). 2. Results A total of 79 participants completed the study. Descriptive characteristics of the participants and scores on all balance tests and questionnaires are provided in Table 1. All three BESTest measures showed moderate association with the ABC scale (r = 0.62–0.675, p < 0.001), and the TUG (r = 0.60 to 0.68, p < 0.001), and a fair association with the PASE (r = 0.33– 0.40, p < 0.005) (Table 2). The BESTest and the mini-BESTest had a moderate association with the SLS (r = 0.67 and 0.68 respectively, p < 0.001), whereas the briefBESTest had a high association with the SLS (r = 0.77, p < 0.001) (Table 2). Twenty-six of 78 participants (33.3%) were classified as fallers as they reported having one or more falls within the previous year.
3
Table 1 Participant characteristics. Measure (units)
Mean (SD)
Min, Max
Age (y) Height (m) Weight (kg) BMI (kg/m2) BESTest (0–100%) mini-BESTest (0–28 points) (n = 76) briefBESTest (0–24 points) ABC (0–100%) PASE (0–400 points) TUG (s) (n = 75) SLS (s) (n = 74)
68.7 1.7 71.5 25.5 88.1 25.9 19.3 94.0 161.9 8.7 22.8
50, 87 1.5, 1.9 50.5, 116.0 18.6, 40.1 55.6, 99.1 11.0, 32.0 7.0, 24.0 69.4, 100 6.4, 447.0 5.4, 14.9 3.0, 30.0
(10.57) (0.1) (13.9) (4.0) (8.8) (4.7) (4.2) (6.5) (83.2) (1.9) (10.2)
n = 79 unless otherwise specified. BMI = body mass index; BESTest = Balance Evaluation Systems Test; ABC = activitiesspecific balance confidence; PASE = Physical Activity Scale for the Elderly; TUG = Timed Up and Go; SLS = Single Leg Stance. For the BESTest, mini-BESTest, briefBESTest, PASE and SLS, higher scores indicate better performance. For the TUG, lower values indicate better performance.
Table 2 Pearson correlations between BESTest, mini-BESTest and briefBESTest and balance confidence, physical activity level and other measures of balance (p < 0.001 unless otherwise stated). BESTest ABC PASE TUG SLS
0.67 0.39 0.68 0.67
mini-BESTest
briefBESTest
0.62 0.33* 0.66 0.68
0.66 0.40 0.60 0.77
ABC = Activities-Specific Balance Confidence; PASE = Physical Activity Scale for the Elderly; TUG = Timed Up and Go; SLS = Single Leg Stance. * p = 0.004.
Table 3 Difference in test scores between fallers and non-fallers; mean (SD). Measure (units)
Fallers
Non-fallers
p
BESTest (0–100%) mini-BESTest (0–28 points) briefBESTest (0–24 points) ABC (0–100%) TUG (s) SLS (s)
85.8 (9.8) 25.2 (5.2)
89.2 (8.2) 26.3 (4.5)
0.054 0.17
3.4 ( 0.77–7.62) 1.1 ( 1.18–3.46)
18.4 (4.6)
19.8 (4.0)
0.08
1.4 ( 0.58–3.43)
91.6 (6.5) 9.0 (2.2) 20.8 (11.0)
95.1 (6.3) 8.5 (1.7) 24.0 (9.6)
0.013 0.12 0.10
3.5 (0.43–6.48) 0.5 ( 1.43–0.37) 3.2 ( 1.75–8.15)
Mean difference (95% CI)
BESTest = Balance Evaluation Systems Test; ABC = activities-specific balance confidence; TUG = Timed Up and Go; SLS = Single Leg Stance. For the BESTest, miniBESTest, briefBESTest, and SLS higher scores indicate better performance. For the TUG, lower values indicate better performance.
Table 4 Difference in test scores between participants at high risk and low risk of falls; mean (SD). Measure (units)
High risk
Low risk
p
Mean difference (95% CI)
BESTest (0–100%) mini-BESTest (0–28 points) briefBESTest (0–24 points) ABC (0–100%) TUG (s) SLS (s)
80.0 (12.7) 22.2 (6.1)
89.8 (6.7) 26.8 (4.0)
0.007 0.009
9.8 (2.42–17.36) 4.6 (0.94–8.18)
16.2 (5.5)
20.0 (3.6)
0.014
3.8 (0.48–7.03)
89.6 (6.1) 9.9 (2.5) 14.8 (10.4)
94.9 (6.2) 8.5 (1.6) 24.8 (9.2)
0.003 0.026 <0.001
5.3 (1.64–8.89) 1.4 ( 2.95–0.02) 10.0 (4.32–15.75)
BESTest = Balance Evaluations Systems Test; ABC = Activities-Specific Balance Confidence; TUG = Timed Up and Go; SLS = Single Leg Stance. For the BESTest, mini-BESTest, briefBESTest and SLS, higher scores indicate better performance. For the TUG, lower values indicate better performance.
Please cite this article in press as: O’Hoski S, et al. Construct validity of the BESTest, mini-BESTest and briefBESTest in adults aged 50 years and older. Gait Posture (2015), http://dx.doi.org/10.1016/j.gaitpost.2015.06.006
G Model
GAIPOS-4507; No. of Pages 5 4
S. O’Hoski et al. / Gait & Posture xxx (2015) xxx–xxx
The difference in balance scores between fallers and non-fallers was borderline statistically significant for the BESTest (p = 0.054) but not for the mini-BESTest (p = 0.17) or briefBESTest (p = 0.08). Fallers had significantly lower ABC scores (p = 0.013) but there was no significant difference in their SLS (p = 0.12) or TUG performance (p = 0.10) (Table 3). Sixty-five of 79 participants (82.3%) were classified as low risk for falls based on their EFST score. The 16 people who were classified as high risk (2 out of 5 points) had significantly lower scores on the BESTest, mini-BESTest, and briefBESTest. These participants also had significantly lower scores on the ABC, TUG and SLS tests (see Table 4). 3. Discussion This is the first study to examine the construct validity of the BESTest, mini-BESTest and briefBESTest in a community-based sample of middle-aged and elderly people. Our results showed that all three BESTest measures had moderate to high associations with other balance measures and fair associations with physical activity, with all three BESTest measures performing similarly. In addition, in this sample, we found that the BESTest and its abbreviated versions performed similarly to other commonly used balance measures for discriminating between those at low and high risk of falls. These findings support the construct validity of the BESTest, mini-BESTest and briefBESTest for assessing balance in adults aged over 50 years. Ultimately, the selection of the most appropriate BESTest version should be based on the purpose of and time available for the assessment. In accordance with our primary hypothesis, all three BESTest scores showed higher correlations with balance confidence and with other measures of balance and lower correlations with physical activity. A moderate association between the ABC scale and other measures of balance has been shown previously in communitydwelling elderly (r = 0.54) [21]. Our results are also in line with research describing the relationship between the BESTest and the ABC scale in a variety of clinical populations (r = 0.69) [4] and with work showing a similar association (r = 0.66) between scores on the mini-BESTest and ABC scale in people with Parkinson’s disease [22]. Two of the most commonly used balance assessments in physical therapy are the TUG and the SLS [12]. We found all three BESTest measures to correlate moderately with the TUG (r = 0.60 to 0.68) which falls within the range of correlations previously demonstrated between other performance-based balance tests and the TUG in older adults (r = 0.69 to 0.83) [23]. We also found moderate to high correlations between the BESTest measures and the SLS (r = 0.67–0.77), consistent with previous work showing a similar associations between the SLS and another multi-item dynamic balance test in older adults with chronic stroke (r = 0.65– 0.79) [24]. The full BESTest and brief-BESTest have not previously been studied in relation to the TUG or SLS, but the mini-BESTest has been shown to be associated with both in individuals with chronic stroke [25]. Our findings for the relationship between BESTest scores and physical activity (r = 0.33–0.40) were also similar to previous reports of the association between the PASE and measures of balance in older adults (r = 0.33–0.38) [16,21] and in patients with COPD, in which the PASE showed similar correlations with the BESTest (r = 0.40) [6]. Our second hypothesis was that participants with a selfreported fall history would score significantly lower on all three of the BESTest measures. Although there was a trend (p = 0.054) for the BESTest to distinguish between fallers and non-fallers, neither of the abbreviated versions discriminated between groups. These results are in contrast with a prior study in which all three versions of the BESTest discriminated between fallers and non-fallers in people with and without a neurological diagnosis [9], and with a study in which the full BESTest and mini-BESTest discriminated
between fallers and non-fallers with Parkinson’s disease [7]. However, similar to the BESTest results, in this sample of adults over 50 years, the TUG and SLS also failed to discriminate between fallers and non-fallers. These results are consistent with a recent meta-analysis that showed that the TUG score had limited ability to discriminate between prospective fallers and non-fallers [19] but differ from a previous study in which the SLS characterized fallers in community-dwelling elderly [14]. There are several reasons that our findings may not be consistent with previous literature. First, participants in this study were likely higher functioning than in previous work. Indeed, mean performance on the balance measures in our study was better than that reported previously for fallers and non-fallers in studies of older adults (mean TUG 8.7 versus 11.4–14.6 s and mean SLS 22.8 versus 8.3 s) [14,23]. In addition, the fall rate in our sample was lower than in previous work [23] and the self-reported physical activity level was higher [17]. Second, although the mean age of our sample was 68.7 years, it is possible that inclusion of middle-aged participants (age 50–60 years (n = 20, 25.3%)) affected the results since most prior work included adults over the age of 65. Third, it is difficult to compare to previous work as a result of differing definitions of fallers. We classified any participant with self-reported falls in the previous year as a faller, but other work used time periods such as 3 months [9] or 6 months [7] or classified fallers as those reporting two or more falls within the timeframe [7]. Taking these factors into consideration, it is likely that the fallers in our study actually had better balance than participants in previous studies. It has been previously noted that those who are more physically active (and therefore are likely to have better balance) [26] may in fact fall due to increased exposure to extrinsic fall risk factors rather than deficits in balance [2,27]. Therefore, in higher functioning adults, individuals who fall may not be characterized by deficits in balance. Our findings showing that poorer performance on all three versions of the BESTest – along with the ABC, TUG, and SLS – characterized the smaller proportion of participants (20%) classified as having high risk of falls based on their EFST score supports this hypothesis. The EFST includes not only fall history but also self-reported near falls and an objective evaluation of gait to identify lower-functioning patients at high risk of falling because of their physical functional deficits [20]. Thus despite negative findings related to identifying those with a history of one or more falls, our results support the known-groups validity of the BESTest and abbreviated versions with respect to fall risk categorized using more stringent criteria. 4. Study limitations Our study has three main limitations. First, our findings are only generalizable to Canadians between 50 and 89 years of age living in urban areas. A second limitation is that participants were classified as fallers based on self-report and their responses were subject to recall bias. Third, since all balance tests were scored concurrently based on a single performance of each task, it is not surprising that the tests performed similarly in our analyses. However, given this consistency in the scoring of the tasks, we can be confident that any differences in test performance are due to the inclusion of different balance tasks as opposed to differences in performance or scoring. 5. Conclusion While our results support the construct validity of the BESTest, mini-BESTest and briefBESTest in adults aged over 50 years, further research is required to determine the predictive validity and responsiveness to change of these tests. Given that the BESTest, mini-BESTest and briefBESTest have comparable construct validity, the choice of test should be based on the time allowed for administration and the purpose of the assessment. Use of the full
Please cite this article in press as: O’Hoski S, et al. Construct validity of the BESTest, mini-BESTest and briefBESTest in adults aged 50 years and older. Gait Posture (2015), http://dx.doi.org/10.1016/j.gaitpost.2015.06.006
G Model
GAIPOS-4507; No. of Pages 5 S. O’Hoski et al. / Gait & Posture xxx (2015) xxx–xxx
BESTest is appropriate when there is time to complete a comprehensive balance assessment; when shorter tests are required, the briefBESTest may be preferred as a tool to guide treatment planning while the mini-BESTest may be more suitable as a dynamic balance screening tool. Acknowledgments We would like to acknowledge the assistance of Bonnie Winship, Lauren Herridge and Taimoor Agha with participant recruitment and data collection. Marla Beauchamp is supported by a fellowship from the Canadian Institutes of Health Research (grant #201202). Conflict of interest The authors have no conflicts of interest to declare. References [1] SMARTRISK. The economic burden of injury in Canada. Toronto, ON: SMARTRISK; 2009. [2] Delbaere K, Close JC, Heim J, Sachdev PS, Brodaty H, Slavin MJ, et al. A multifactorial approach to understanding fall risk in older people. J Am Geriatr Soc 2010;58(9):1679–85. [3] Horak FB. Postural orientation and equilibrium: what do we need to know about neural control of balance to prevent falls? Age Ageing 2006;35(Suppl. 2):7–11. [4] Horak FB, Wrisley DM, Frank J. The balance evaluation systems test (BESTest) to differentiate balance deficits. Phys Ther 2009;89(5):484–98. [5] Sibley KM, Beauchamp MK, Van Ooteghem K, Straus SE, Jaglal SB. Using the systems framework for postural control to analyze the components of balance evaluated in standardized balance measures: a scoping review. Arch Phys Med Rehabil 2015;96(1):122–32. [6] Beauchamp MK, Sibley KM, Lakhani B, Romano J, Mathur S, Goldstein RS, et al. Impairments in systems underlying control of balance in COPD. Chest 2012;141(6):1496–503. [7] Leddy AL, Crowner BE, Earhart GM. Utility of the mini-BESTest, BESTest, and BESTest sections for balance assessments in individuals with Parkinson disease. J Neurol Phys Ther 2011;35(2):90–7. [8] O’Hoski S, Winship B, Herridge L, Agha T, Brooks D, Beauchamp MK, et al. Increasing the clinical utility of the BESTest, mini-BESTest, and briefBESTest: normative values in Canadian adults who are healthy and aged 50 years or older. Phys Ther 2014;94(Mar (3)):334–42. [9] Padgett PK, Jacobs JV, Kasser SL. Is the BESTest at its best? A suggested brief version based on interrater reliability, validity, internal consistency, and theoretical construct. Phys Ther 2012;92(9):1197–207.
5
[10] Franchignoni F, Horak F, Godi M, Nardone A, Giordano A. Using psychometric techniques to improve the balance evaluation systems test: the mini-BESTest. J Rehabil Med 2010;42(4):323–31. [11] Duncan RP, Leddy AL, Cavanaugh JT, Dibble LE, Ellis TD, Ford MP, et al. Comparative utility of the BESTest, mini-BESTest, and brief-BESTest for predicting falls in individuals with Parksinson disease: a cohort study. Phys Ther 2013;93(Apr (4)):542–50. [12] Sibley KM, Straus SE, Inness EL, Salbach NM, Jaglal SB. Balance assessment practices and use of standardized balance measures among Ontario physical therapists. Phys Ther 2011;91(11):1583–91. [13] Powell LE, Myers AM. The activities-specific balance confidence (ABC) scale. J Gerontol Ser A-Biol Sci Med Sci 1995;50(1):M28–34. [14] Lin MR, Hwang HF, Hu MH, Wu HD, Wang YW, Huang FC. Psychometric comparisons of the timed up and go, one-leg stand, functional reach, and Tinetti balance measures in community-dwelling older people. J Am Geriatr Soc 2004;52(Aug (8)):1343–8. [15] Lajoie Y, Gallagher SP. Predicting falls within the elderly community: comparison of postural sway, reaction time, the Berg balance scale and the Activities-specific Balance Confidence (ABC) scale for comparing fallers and non-fallers. Arch Gerontol Geriatr 2004;38(Jan–Feb (1)):11–26. [16] Washburn R, Smith K, Jette A, Janney C. The physical activity scale for the elderly (PASE): development and evaluation. J Clin Epidemiol 1993;46(2): 153–62. [17] Washburn RA, Ficker JL. Physical activity scale for the elderly (PASE): the relationship with activity measured by a portable accelerometer. J Sports Med Phys Fitness 1999;39(4):336–40. [18] Podsiadlo D, Richardson S. The timed up & go: a test of basic functional mobility for frail elderly persons. J Am Geriatr Soc 1991;39(2):142–8. [19] Barry E, Galvin R, Keogh C, Horgan F, Fahey T. Is the timed up and go test a useful predictor of risk of falls in community dwelling older adults: a systematic review and meta-analysis. BMC Geriatr 2014;14:14. [20] Cwikel J, Fried A, Biderman A, Galinsky D. Validation of a fall-risk screening test, the elderly fall screening test (EFST), for community-dwelling elderly. Disabil Rehabil 1998;20(5):161–7. [21] Nemmers TM, Williams Miller J. Factors influencing balance in healthy community-dwelling women age 60 and older. J Geriatr Phys Ther 2008;31(3):93– 100. [22] McNeely ME, Duncan RP, Earhart GM. Medication improves balance and complex gait performance in Parkinson disease. Gait Posture 2012;36(1): 144–8. [23] Gonc¸alves DFF, Ricii NA, Coimbra AMV. Functional balance among older adults from the community: a comparison of their history of falls. Rev Bras Fisioter 2009;13(4):316–23. [24] Flansbjer UB, Blom J, Broga˚rdh C. The reproducibility of Berg Balance Scale and the single-leg stance in chronic stroke and the relationship between the two tests. PMR 2012;4(Mar (3)):165–70. [25] Tsang CS, Liao LR, Chung RC, Pang MY. Psychometric properties of the MiniBalance Evaluation Systems Test (Mini-BESTest) in community-dwelling individuals with chronic stroke. Phys Ther 2013;93(Aug (8)):1102–15. [26] Howe TE, Rochester L, Jackson A, Banks PMH, Blair VA. Exercise for improving balance in older people. Cochrane Database Syst Rev 2007;4:CD004963. [27] O’Loughlin JL, Robitaille Y, Boivin J-F, Suissa S. Incidence of and risk factors for falls and injurious falls among the community-dwelling elderly. Am J Epidemiol 1993;137(3):342–54.
Please cite this article in press as: O’Hoski S, et al. Construct validity of the BESTest, mini-BESTest and briefBESTest in adults aged 50 years and older. Gait Posture (2015), http://dx.doi.org/10.1016/j.gaitpost.2015.06.006
GAIPOS-4507; No. of Pages 5 Gait & Posture xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Construct validity of the BESTest, mini-BESTest and briefBESTest in adults aged 50 years and older Sachi O’Hoski a, Kathryn M. Sibley b,c, Dina Brooks a,b,c, Marla K. Beauchamp a,d,* a
West Park Healthcare Centre, Respiratory Medicine, 82 Buttonwood Avenue, Toronto, ON, M6M 2J5, Canada University of Toronto, Department of Physical Therapy, Faculty of Medicine, 500 University Avenue, Toronto, ON M5G 1V7, Canada c Toronto Rehabilitation Institute – University Health Network, 550 University Avenue, Toronto, ON M5G 2A2, Canada d Department of Physical Medicine and Rehabilitation, Harvard Medical School, Spaulding Rehabilitation Hospital, Cambridge, MA, USA b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 22 November 2014 Received in revised form 5 June 2015 Accepted 15 June 2015
Background: The Balance Evaluation Systems Test (BESTest) and its two abbreviated versions (miniBESTest and briefBESTest) are functional balance tools that have yet to be validated in middle aged and elderly people living in the community. Objective: Determine the construct validity of the three BESTest versions by comparing them with commonly-used measures of balance, balance confidence and physical activity, and examining their ability to discriminate between groups with respect to falls and fall risk. Methods: This was a secondary analysis of data from 79 adults (mean age 68.7 10.57 years). Pearson correlation coefficients were used to examine the relationships between each BESTest measure and the Activities-Specific Balance Confidence (ABC) scale, the Physical Activity Scale for the Elderly (PASE), the Timed Up and Go (TUG) and the Single Leg Stance (SLS) test. Independent t-tests were used to examine differences in balance between fallers (1 fall in previous year) and non-fallers and individuals classified at low versus high fall risk using the Elderly Falls Screening Test (EFST). Results: The BESTest measures showed moderate associations with the ABC scale and TUG (r = 0.62–0.67 and 0.60 to 0.68 respectively), fair associations (r = 0.33–0.40) with the PASE and moderate to high associations (r = 0.67–0.77) with the SLS. Fallers showed a trend (p = 0.054) for lower scores on the original BESTest, and people at high risk for falls had significantly lower scores on all BESTest versions. Conclusions: These findings support the construct validity of the BESTest, mini-BESTest and briefBESTest in adults over 50 years old. ß 2015 Published by Elsevier B.V.
Keywords: Postural balance Geriatric assessment Outcomes assessment Aged
Falls in later adulthood are a significant public health problem [1]. While numerous fall risk factors have been identified, balance impairment is recognized as one of the most important modifiable risk factors for falls [2]. The ability to maintain balance is a complex skill that requires the integration of information from multiple physiological systems [3], many of which are impaired with age [2]. Clinical measures of balance with sound psychometric properties are needed to screen for balance problems in order to guide treatment and inform fall prevention strategies. The Balance Evaluation Systems Test (BESTest) is a clinical balance tool that targets six subsystems of postural control (see Box 1) in order to identify the underlying impairments contributing
* Corresponding author at: Department of Physical Medicine and Rehabilitation, Harvard Medical School, Spaulding Outpatient Center Cambridge, 1575 Cambridge St., Cambridge, MA 02138, USA. Tel.: +1 617 952 6954; fax: +1 617 952 6965. E-mail addresses: [email protected], [email protected] (M.K. Beauchamp).
to dysfunctional balance [4]. In a recent review, it was identified as the only standardized balance measure that evaluates all components of balance consistent with established conceptual models [5]. The BESTest has been used in a variety of populations [4,6] and has been shown to have strong psychometric properties in people with Parkinson’s disease [7]. However, few studies have reported on its use in a general population of middle-aged and older adults [4,6,8]. Two abbreviated versions of the BESTest (mini-BESTest and briefBESTest) have been developed as alternatives to the original test. These abbreviated versions take 10–20 min to complete [9,10], considerably less than the 30–60 min to administer the BESTest [4,9,10], time that is not feasible in many clinical settings. The miniBESTest was developed using factor analysis to identify the items of the BESTest that represented dynamic balance. Rasch analysis was then used to improve the rating categories and eliminate some items [10]. The result was a 14 item test of dynamic balance [10]. The miniBESTest has been used in several clinical populations [10] and evidence for its psychometric properties has been shown in people
http://dx.doi.org/10.1016/j.gaitpost.2015.06.006 0966-6362/ß 2015 Published by Elsevier B.V.
Please cite this article in press as: O’Hoski S, et al. Construct validity of the BESTest, mini-BESTest and briefBESTest in adults aged 50 years and older. Gait Posture (2015), http://dx.doi.org/10.1016/j.gaitpost.2015.06.006
G Model
GAIPOS-4507; No. of Pages 5 S. O’Hoski et al. / Gait & Posture xxx (2015) xxx–xxx
2
Box 1. Description of categories and items of the Balance Evaluation Systems Test. Category
Description and examples of tasks
Biomechanical constraints
Evaluates constraints on standing balance including postural alignment, ankle range of motion and hip strength Evaluates ability to move the body over its base of support by leaning forward and laterally and evaluates ability to return to gravitational vertical Evaluates active movement of the center of mass in anticipation of performing sit to stand, Single Leg Stance, and stair tap Evaluates in-place and compensatory stepping responses to external perturbations from the front, side and back Evaluates increases in postural sway under different sensory conditions such as standing on flat ground or foam with eyes open or closed Evaluates stability while walking under different conditions such as changing speed, looking from side to side and stepping over an obstacle
Stability limits/verticality
Anticipatory postural adjustments Postural responses
Sensory orientation
Stability in gait
with Parkinson’s disease [7]. Unlike the mini-BESTest, the briefBESTest maintains the original theoretical basis of the BESTest; it was created using item-total correlations to identify the most representative item from each subsection of the BESTest [9]. The brief-BESTest has similar inter-rater reliability to the BESTest and mini-BESTest and is better able to discriminate between fallers and non-fallers with and without a neurological diagnosis [9]. The construct validity of the BESTest, mini-BESTest and briefBESTest has not yet been examined in a general community-based sample. Therefore, our primary objective was to determine the associations between the BESTest and its abbreviated versions and measures of other theoretically related constructs (i.e., convergent validity) including measures of balance, balance confidence and physical activity in adults aged over 50 years living in the community. We hypothesized that all three BESTest scores would have high correlations with self-reported balance confidence and other performance-based measures of balance and moderate correlations with physical activity. In addition, while the BESTest, mini-BESTest and briefBESTest have been shown to discriminate between fallers and non-fallers with Parkinson’s disease [7,11], this has not been examined in those without neurological conditions. Therefore, our secondary objective was to determine the ability of the BESTest, mini-BESTest and briefBESTest to discriminate between groups with respect to fall history and risk of falls and to compare the discriminative ability of the BESTest measures with that of commonly used balance measures [12]; the self-reported Activities-Specific Balance Confidence (ABC) scale [13], and the Timed Upand-Go (TUG) test [14] and Single Leg Stance (SLS) test [14]. We hypothesized that fallers and those classified as being at high risk of falls would score significantly lower on all balance measures. 1. Methods This was a secondary analysis of a previous study; the methods have been described in detail elsewhere [8]. Briefly, the study protocol was approved by the University Research Ethics Board and written informed consent was obtained at the beginning of each data collection session. Individuals who met the following criteria were recruited: (1) age between 50 and 89 years, (2) residing independently in the community, (3) able to understand and follow three-step instructions, (4) able to walk 6 m independently
(without a gait aid). Participants were excluded if they reported: (1) a history of dizziness or fainting, (2) the presence of any cardiorespiratory, neurological or musculoskeletal condition that severely affected their balance (e.g., chronic obstructive pulmonary disease, stroke, recent hip or knee replacement), or (3) the use of medication(s) that they felt caused dizziness or affected their balance (e.g., antidepressants). Data were collected between January and July 2012 by raters trained in the administration of the BESTest. Demographic data (sex, age, height, weight) and all written questionnaires were collected before administering the BESTest. Two testers were present for each testing session; one to observe and score the items and one to supervise for safety. Scoring of the abbreviated tests was completed after the data collection session based on performance of the original BESTest tasks. 1.1. Outcome measures 1.1.1. Balance Evaluation Systems Test (BESTest) The BESTest is a 36 item test comprised of six subsections that contribute to postural control (see Box 1) [4]. Each item is scored from 0 to 3 points based on time or performance criteria resulting in a total possible score of 108 points, which is converted to a percentage score. Higher scores indicate better balance. 1.1.2. mini-BESTest The mini-BESTest is a 14-item test of dynamic balance that includes tasks from the BESTest subsystems ‘anticipatory postural adjustments’, ‘postural responses’, ‘sensory orientation’ and ‘stability in gait’ [10]. Each item is scored from 0 to 2 points resulting in a total score out of 28 points with higher scores indicating better balance. 1.1.3. briefBESTest The briefBESTest is an eight-item test comprised of one item from each subsection of the BESTest, with two items (SLS and functional reach forward) scored bilaterally [9]. Each item is scored from 0 to 3 points with a total possible score out of 24 points (higher scores for better balance). 1.1.4. Activities-Specific Balance Confidence (ABC) scale The ABC scale is a 16-item self-report questionnaire requiring individuals to indicate their confidence in completing progressively difficult tasks without becoming unsteady or losing their balance (0% = no confidence; 100% = completely confident) [13]. The percentage assigned for each task is summed and divided by 16 in order to obtain the overall score. The ABC scale has good test–retest reliability, convergent and criterion validity [13] and the ability to discriminate between fallers and non-fallers in community-dwelling elderly (cut-off score 67%) [15]. 1.1.5. Physical Activity Scale for the Elderly (PASE) The PASE consists of 26 questions that assess an individual’s physical activity level over the previous 7 days [16]. Subscale categories include leisure time, household and occupational physical activity. The frequency (never, seldom, sometimes, often) and duration (hours) of physical activity is recorded. Responses to individual items are weighted and summed to calculate the subscores and total score, which can range from 0 to 400 points. A high score is indicative of a high level of physical activity. The PASE has good test–retest reliability [16] and good construct validity when scores are compared to measures of static balance [16] and accelerometer data [17]. 1.1.6. Timed Up and Go (TUG) test Participants are instructed to stand up from a chair, walk 3 m at their usual walking speed, turn 1808, walk back to the chair and sit
Please cite this article in press as: O’Hoski S, et al. Construct validity of the BESTest, mini-BESTest and briefBESTest in adults aged 50 years and older. Gait Posture (2015), http://dx.doi.org/10.1016/j.gaitpost.2015.06.006
G Model
GAIPOS-4507; No. of Pages 5 S. O’Hoski et al. / Gait & Posture xxx (2015) xxx–xxx
down [18]. The entire task is timed and a lower time to completion indicates better mobility [18]. The TUG has been shown to have excellent intra- and inter-rater reliability and good convergent validity with other measures of balance [14]. While the traditional cut-off (13.5 s) has good specificity for ‘‘ruling in’’ falls, a recent metaanalysis noted it had low sensitivity and limited ability to discriminate between prospective fallers and non-fallers in community-dwelling older adults (area under the curve (AUC) = 0.57) [19]. 1.1.7. Single Leg Stance (SLS) test For this task, participants are asked to place their hands on their hips, lift one foot off the floor and hold that position for as long as possible. Higher times indicate better balance performance. The best time on each leg was recorded but only the highest time achieved was used for data analysis. SLS time has been shown to have excellent test–retest reliability [14] and to be able to discriminate between fallers and non-fallers in community dwelling elderly (AUC = 0.64) [14]. 1.1.8. Elderly Falls Screening Test (EFST) The EFST is a five-item test that combines self-reported fall questions with a gait evaluation (quality and speed) to determine fall risk [20]. Respondents are asked to indicate whether or not they have experienced a fall in the last year, defined as finding themselves ‘‘suddenly on the ground, without intending to get there, after [they] were in either a lying, sitting or standing position’’. For this study we classified ‘fallers’ as those who responded ‘‘yes’’ and ‘non-fallers’ as those who responded ‘‘no’’. The EFST total score ranges between 0 (low fall risk) and 5 (high fall risk), with participants receiving one point for each of the following: (1) two or more falls in the past year, (2) any injury from a fall, (3) occasional or frequent near falls, (4) taking longer than 10 s to walk 5 m, and (5) uneven, shuffling, wide-based, or unsteady gait. Those with total score of 2 points are considered at high-risk of future falls [20]. The EFST has good criterion and predictive validity for falls in the elderly [20]. 1.2. Data analysis Descriptive statistics (mean, SD, range) were calculated for age, height, weight, body mass index and each of the measures. Pearson correlation coefficients were calculated to examine the relationship between the BESTest, mini-BESTest, and briefBESTest, and the ABC, PASE, TUG, and SLS. By convention, a correlation of 0.00–0.25 was interpreted to indicate little or no relationship, 0.25–0.5 as a fair relationship, 0.5–0.75 as moderate, and above 0.75 as a very good to excellent relationship. Independent t-tests were used to examine differences between fallers versus non-fallers and individuals at low versus high fall risk based on the EFST. All analyses were conducted with SPSS (version 19.0 for Windows; SPSS Inc.; Chicago, United States). 2. Results A total of 79 participants completed the study. Descriptive characteristics of the participants and scores on all balance tests and questionnaires are provided in Table 1. All three BESTest measures showed moderate association with the ABC scale (r = 0.62–0.675, p < 0.001), and the TUG (r = 0.60 to 0.68, p < 0.001), and a fair association with the PASE (r = 0.33– 0.40, p < 0.005) (Table 2). The BESTest and the mini-BESTest had a moderate association with the SLS (r = 0.67 and 0.68 respectively, p < 0.001), whereas the briefBESTest had a high association with the SLS (r = 0.77, p < 0.001) (Table 2). Twenty-six of 78 participants (33.3%) were classified as fallers as they reported having one or more falls within the previous year.
3
Table 1 Participant characteristics. Measure (units)
Mean (SD)
Min, Max
Age (y) Height (m) Weight (kg) BMI (kg/m2) BESTest (0–100%) mini-BESTest (0–28 points) (n = 76) briefBESTest (0–24 points) ABC (0–100%) PASE (0–400 points) TUG (s) (n = 75) SLS (s) (n = 74)
68.7 1.7 71.5 25.5 88.1 25.9 19.3 94.0 161.9 8.7 22.8
50, 87 1.5, 1.9 50.5, 116.0 18.6, 40.1 55.6, 99.1 11.0, 32.0 7.0, 24.0 69.4, 100 6.4, 447.0 5.4, 14.9 3.0, 30.0
(10.57) (0.1) (13.9) (4.0) (8.8) (4.7) (4.2) (6.5) (83.2) (1.9) (10.2)
n = 79 unless otherwise specified. BMI = body mass index; BESTest = Balance Evaluation Systems Test; ABC = activitiesspecific balance confidence; PASE = Physical Activity Scale for the Elderly; TUG = Timed Up and Go; SLS = Single Leg Stance. For the BESTest, mini-BESTest, briefBESTest, PASE and SLS, higher scores indicate better performance. For the TUG, lower values indicate better performance.
Table 2 Pearson correlations between BESTest, mini-BESTest and briefBESTest and balance confidence, physical activity level and other measures of balance (p < 0.001 unless otherwise stated). BESTest ABC PASE TUG SLS
0.67 0.39 0.68 0.67
mini-BESTest
briefBESTest
0.62 0.33* 0.66 0.68
0.66 0.40 0.60 0.77
ABC = Activities-Specific Balance Confidence; PASE = Physical Activity Scale for the Elderly; TUG = Timed Up and Go; SLS = Single Leg Stance. * p = 0.004.
Table 3 Difference in test scores between fallers and non-fallers; mean (SD). Measure (units)
Fallers
Non-fallers
p
BESTest (0–100%) mini-BESTest (0–28 points) briefBESTest (0–24 points) ABC (0–100%) TUG (s) SLS (s)
85.8 (9.8) 25.2 (5.2)
89.2 (8.2) 26.3 (4.5)
0.054 0.17
3.4 ( 0.77–7.62) 1.1 ( 1.18–3.46)
18.4 (4.6)
19.8 (4.0)
0.08
1.4 ( 0.58–3.43)
91.6 (6.5) 9.0 (2.2) 20.8 (11.0)
95.1 (6.3) 8.5 (1.7) 24.0 (9.6)
0.013 0.12 0.10
3.5 (0.43–6.48) 0.5 ( 1.43–0.37) 3.2 ( 1.75–8.15)
Mean difference (95% CI)
BESTest = Balance Evaluation Systems Test; ABC = activities-specific balance confidence; TUG = Timed Up and Go; SLS = Single Leg Stance. For the BESTest, miniBESTest, briefBESTest, and SLS higher scores indicate better performance. For the TUG, lower values indicate better performance.
Table 4 Difference in test scores between participants at high risk and low risk of falls; mean (SD). Measure (units)
High risk
Low risk
p
Mean difference (95% CI)
BESTest (0–100%) mini-BESTest (0–28 points) briefBESTest (0–24 points) ABC (0–100%) TUG (s) SLS (s)
80.0 (12.7) 22.2 (6.1)
89.8 (6.7) 26.8 (4.0)
0.007 0.009
9.8 (2.42–17.36) 4.6 (0.94–8.18)
16.2 (5.5)
20.0 (3.6)
0.014
3.8 (0.48–7.03)
89.6 (6.1) 9.9 (2.5) 14.8 (10.4)
94.9 (6.2) 8.5 (1.6) 24.8 (9.2)
0.003 0.026 <0.001
5.3 (1.64–8.89) 1.4 ( 2.95–0.02) 10.0 (4.32–15.75)
BESTest = Balance Evaluations Systems Test; ABC = Activities-Specific Balance Confidence; TUG = Timed Up and Go; SLS = Single Leg Stance. For the BESTest, mini-BESTest, briefBESTest and SLS, higher scores indicate better performance. For the TUG, lower values indicate better performance.
Please cite this article in press as: O’Hoski S, et al. Construct validity of the BESTest, mini-BESTest and briefBESTest in adults aged 50 years and older. Gait Posture (2015), http://dx.doi.org/10.1016/j.gaitpost.2015.06.006
G Model
GAIPOS-4507; No. of Pages 5 4
S. O’Hoski et al. / Gait & Posture xxx (2015) xxx–xxx
The difference in balance scores between fallers and non-fallers was borderline statistically significant for the BESTest (p = 0.054) but not for the mini-BESTest (p = 0.17) or briefBESTest (p = 0.08). Fallers had significantly lower ABC scores (p = 0.013) but there was no significant difference in their SLS (p = 0.12) or TUG performance (p = 0.10) (Table 3). Sixty-five of 79 participants (82.3%) were classified as low risk for falls based on their EFST score. The 16 people who were classified as high risk (2 out of 5 points) had significantly lower scores on the BESTest, mini-BESTest, and briefBESTest. These participants also had significantly lower scores on the ABC, TUG and SLS tests (see Table 4). 3. Discussion This is the first study to examine the construct validity of the BESTest, mini-BESTest and briefBESTest in a community-based sample of middle-aged and elderly people. Our results showed that all three BESTest measures had moderate to high associations with other balance measures and fair associations with physical activity, with all three BESTest measures performing similarly. In addition, in this sample, we found that the BESTest and its abbreviated versions performed similarly to other commonly used balance measures for discriminating between those at low and high risk of falls. These findings support the construct validity of the BESTest, mini-BESTest and briefBESTest for assessing balance in adults aged over 50 years. Ultimately, the selection of the most appropriate BESTest version should be based on the purpose of and time available for the assessment. In accordance with our primary hypothesis, all three BESTest scores showed higher correlations with balance confidence and with other measures of balance and lower correlations with physical activity. A moderate association between the ABC scale and other measures of balance has been shown previously in communitydwelling elderly (r = 0.54) [21]. Our results are also in line with research describing the relationship between the BESTest and the ABC scale in a variety of clinical populations (r = 0.69) [4] and with work showing a similar association (r = 0.66) between scores on the mini-BESTest and ABC scale in people with Parkinson’s disease [22]. Two of the most commonly used balance assessments in physical therapy are the TUG and the SLS [12]. We found all three BESTest measures to correlate moderately with the TUG (r = 0.60 to 0.68) which falls within the range of correlations previously demonstrated between other performance-based balance tests and the TUG in older adults (r = 0.69 to 0.83) [23]. We also found moderate to high correlations between the BESTest measures and the SLS (r = 0.67–0.77), consistent with previous work showing a similar associations between the SLS and another multi-item dynamic balance test in older adults with chronic stroke (r = 0.65– 0.79) [24]. The full BESTest and brief-BESTest have not previously been studied in relation to the TUG or SLS, but the mini-BESTest has been shown to be associated with both in individuals with chronic stroke [25]. Our findings for the relationship between BESTest scores and physical activity (r = 0.33–0.40) were also similar to previous reports of the association between the PASE and measures of balance in older adults (r = 0.33–0.38) [16,21] and in patients with COPD, in which the PASE showed similar correlations with the BESTest (r = 0.40) [6]. Our second hypothesis was that participants with a selfreported fall history would score significantly lower on all three of the BESTest measures. Although there was a trend (p = 0.054) for the BESTest to distinguish between fallers and non-fallers, neither of the abbreviated versions discriminated between groups. These results are in contrast with a prior study in which all three versions of the BESTest discriminated between fallers and non-fallers in people with and without a neurological diagnosis [9], and with a study in which the full BESTest and mini-BESTest discriminated
between fallers and non-fallers with Parkinson’s disease [7]. However, similar to the BESTest results, in this sample of adults over 50 years, the TUG and SLS also failed to discriminate between fallers and non-fallers. These results are consistent with a recent meta-analysis that showed that the TUG score had limited ability to discriminate between prospective fallers and non-fallers [19] but differ from a previous study in which the SLS characterized fallers in community-dwelling elderly [14]. There are several reasons that our findings may not be consistent with previous literature. First, participants in this study were likely higher functioning than in previous work. Indeed, mean performance on the balance measures in our study was better than that reported previously for fallers and non-fallers in studies of older adults (mean TUG 8.7 versus 11.4–14.6 s and mean SLS 22.8 versus 8.3 s) [14,23]. In addition, the fall rate in our sample was lower than in previous work [23] and the self-reported physical activity level was higher [17]. Second, although the mean age of our sample was 68.7 years, it is possible that inclusion of middle-aged participants (age 50–60 years (n = 20, 25.3%)) affected the results since most prior work included adults over the age of 65. Third, it is difficult to compare to previous work as a result of differing definitions of fallers. We classified any participant with self-reported falls in the previous year as a faller, but other work used time periods such as 3 months [9] or 6 months [7] or classified fallers as those reporting two or more falls within the timeframe [7]. Taking these factors into consideration, it is likely that the fallers in our study actually had better balance than participants in previous studies. It has been previously noted that those who are more physically active (and therefore are likely to have better balance) [26] may in fact fall due to increased exposure to extrinsic fall risk factors rather than deficits in balance [2,27]. Therefore, in higher functioning adults, individuals who fall may not be characterized by deficits in balance. Our findings showing that poorer performance on all three versions of the BESTest – along with the ABC, TUG, and SLS – characterized the smaller proportion of participants (20%) classified as having high risk of falls based on their EFST score supports this hypothesis. The EFST includes not only fall history but also self-reported near falls and an objective evaluation of gait to identify lower-functioning patients at high risk of falling because of their physical functional deficits [20]. Thus despite negative findings related to identifying those with a history of one or more falls, our results support the known-groups validity of the BESTest and abbreviated versions with respect to fall risk categorized using more stringent criteria. 4. Study limitations Our study has three main limitations. First, our findings are only generalizable to Canadians between 50 and 89 years of age living in urban areas. A second limitation is that participants were classified as fallers based on self-report and their responses were subject to recall bias. Third, since all balance tests were scored concurrently based on a single performance of each task, it is not surprising that the tests performed similarly in our analyses. However, given this consistency in the scoring of the tasks, we can be confident that any differences in test performance are due to the inclusion of different balance tasks as opposed to differences in performance or scoring. 5. Conclusion While our results support the construct validity of the BESTest, mini-BESTest and briefBESTest in adults aged over 50 years, further research is required to determine the predictive validity and responsiveness to change of these tests. Given that the BESTest, mini-BESTest and briefBESTest have comparable construct validity, the choice of test should be based on the time allowed for administration and the purpose of the assessment. Use of the full
Please cite this article in press as: O’Hoski S, et al. Construct validity of the BESTest, mini-BESTest and briefBESTest in adults aged 50 years and older. Gait Posture (2015), http://dx.doi.org/10.1016/j.gaitpost.2015.06.006
G Model
GAIPOS-4507; No. of Pages 5 S. O’Hoski et al. / Gait & Posture xxx (2015) xxx–xxx
BESTest is appropriate when there is time to complete a comprehensive balance assessment; when shorter tests are required, the briefBESTest may be preferred as a tool to guide treatment planning while the mini-BESTest may be more suitable as a dynamic balance screening tool. Acknowledgments We would like to acknowledge the assistance of Bonnie Winship, Lauren Herridge and Taimoor Agha with participant recruitment and data collection. Marla Beauchamp is supported by a fellowship from the Canadian Institutes of Health Research (grant #201202). Conflict of interest The authors have no conflicts of interest to declare. References [1] SMARTRISK. The economic burden of injury in Canada. Toronto, ON: SMARTRISK; 2009. [2] Delbaere K, Close JC, Heim J, Sachdev PS, Brodaty H, Slavin MJ, et al. A multifactorial approach to understanding fall risk in older people. J Am Geriatr Soc 2010;58(9):1679–85. [3] Horak FB. Postural orientation and equilibrium: what do we need to know about neural control of balance to prevent falls? Age Ageing 2006;35(Suppl. 2):7–11. [4] Horak FB, Wrisley DM, Frank J. The balance evaluation systems test (BESTest) to differentiate balance deficits. Phys Ther 2009;89(5):484–98. [5] Sibley KM, Beauchamp MK, Van Ooteghem K, Straus SE, Jaglal SB. Using the systems framework for postural control to analyze the components of balance evaluated in standardized balance measures: a scoping review. Arch Phys Med Rehabil 2015;96(1):122–32. [6] Beauchamp MK, Sibley KM, Lakhani B, Romano J, Mathur S, Goldstein RS, et al. Impairments in systems underlying control of balance in COPD. Chest 2012;141(6):1496–503. [7] Leddy AL, Crowner BE, Earhart GM. Utility of the mini-BESTest, BESTest, and BESTest sections for balance assessments in individuals with Parkinson disease. J Neurol Phys Ther 2011;35(2):90–7. [8] O’Hoski S, Winship B, Herridge L, Agha T, Brooks D, Beauchamp MK, et al. Increasing the clinical utility of the BESTest, mini-BESTest, and briefBESTest: normative values in Canadian adults who are healthy and aged 50 years or older. Phys Ther 2014;94(Mar (3)):334–42. [9] Padgett PK, Jacobs JV, Kasser SL. Is the BESTest at its best? A suggested brief version based on interrater reliability, validity, internal consistency, and theoretical construct. Phys Ther 2012;92(9):1197–207.
5
[10] Franchignoni F, Horak F, Godi M, Nardone A, Giordano A. Using psychometric techniques to improve the balance evaluation systems test: the mini-BESTest. J Rehabil Med 2010;42(4):323–31. [11] Duncan RP, Leddy AL, Cavanaugh JT, Dibble LE, Ellis TD, Ford MP, et al. Comparative utility of the BESTest, mini-BESTest, and brief-BESTest for predicting falls in individuals with Parksinson disease: a cohort study. Phys Ther 2013;93(Apr (4)):542–50. [12] Sibley KM, Straus SE, Inness EL, Salbach NM, Jaglal SB. Balance assessment practices and use of standardized balance measures among Ontario physical therapists. Phys Ther 2011;91(11):1583–91. [13] Powell LE, Myers AM. The activities-specific balance confidence (ABC) scale. J Gerontol Ser A-Biol Sci Med Sci 1995;50(1):M28–34. [14] Lin MR, Hwang HF, Hu MH, Wu HD, Wang YW, Huang FC. Psychometric comparisons of the timed up and go, one-leg stand, functional reach, and Tinetti balance measures in community-dwelling older people. J Am Geriatr Soc 2004;52(Aug (8)):1343–8. [15] Lajoie Y, Gallagher SP. Predicting falls within the elderly community: comparison of postural sway, reaction time, the Berg balance scale and the Activities-specific Balance Confidence (ABC) scale for comparing fallers and non-fallers. Arch Gerontol Geriatr 2004;38(Jan–Feb (1)):11–26. [16] Washburn R, Smith K, Jette A, Janney C. The physical activity scale for the elderly (PASE): development and evaluation. J Clin Epidemiol 1993;46(2): 153–62. [17] Washburn RA, Ficker JL. Physical activity scale for the elderly (PASE): the relationship with activity measured by a portable accelerometer. J Sports Med Phys Fitness 1999;39(4):336–40. [18] Podsiadlo D, Richardson S. The timed up & go: a test of basic functional mobility for frail elderly persons. J Am Geriatr Soc 1991;39(2):142–8. [19] Barry E, Galvin R, Keogh C, Horgan F, Fahey T. Is the timed up and go test a useful predictor of risk of falls in community dwelling older adults: a systematic review and meta-analysis. BMC Geriatr 2014;14:14. [20] Cwikel J, Fried A, Biderman A, Galinsky D. Validation of a fall-risk screening test, the elderly fall screening test (EFST), for community-dwelling elderly. Disabil Rehabil 1998;20(5):161–7. [21] Nemmers TM, Williams Miller J. Factors influencing balance in healthy community-dwelling women age 60 and older. J Geriatr Phys Ther 2008;31(3):93– 100. [22] McNeely ME, Duncan RP, Earhart GM. Medication improves balance and complex gait performance in Parkinson disease. Gait Posture 2012;36(1): 144–8. [23] Gonc¸alves DFF, Ricii NA, Coimbra AMV. Functional balance among older adults from the community: a comparison of their history of falls. Rev Bras Fisioter 2009;13(4):316–23. [24] Flansbjer UB, Blom J, Broga˚rdh C. The reproducibility of Berg Balance Scale and the single-leg stance in chronic stroke and the relationship between the two tests. PMR 2012;4(Mar (3)):165–70. [25] Tsang CS, Liao LR, Chung RC, Pang MY. Psychometric properties of the MiniBalance Evaluation Systems Test (Mini-BESTest) in community-dwelling individuals with chronic stroke. Phys Ther 2013;93(Aug (8)):1102–15. [26] Howe TE, Rochester L, Jackson A, Banks PMH, Blair VA. Exercise for improving balance in older people. Cochrane Database Syst Rev 2007;4:CD004963. [27] O’Loughlin JL, Robitaille Y, Boivin J-F, Suissa S. Incidence of and risk factors for falls and injurious falls among the community-dwelling elderly. Am J Epidemiol 1993;137(3):342–54.
Please cite this article in press as: O’Hoski S, et al. Construct validity of the BESTest, mini-BESTest and briefBESTest in adults aged 50 years and older. Gait Posture (2015), http://dx.doi.org/10.1016/j.gaitpost.2015.06.006