Reliability and Validity of Ratings of Perceived Difficulty During Performance of Static Standing Balance Exercises
Alsubaie SF, Whitney SL, Furman JM, et al. Reliability and validity of ratings of perceived difficulty during performance of static standing balance exercises. Physical Therapy. 2019;99(10):1381-1393. doi:10.1093/ptj/pzz091Link to Original Article: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6821262/
Key Points
- The study aimed to develop a method for quantifying the perceived intensity of standing balance exercises.
- Participants performed 2 sets of 24 randomized static standing exercises with different factors such as surface, vision, stance, and head movement.
- Postural sway was measured using an inertial measurement unit, and ratings of perceived difficulty (RPD) were recorded using numerical and qualitative scales.
- Both RPD scales were found to be associated with postural sway measures, suggesting that they can be used as a proxy measure for perceived intensity.
- The test-retest reliability of the RPD scales varied across different balance exercises, with the highest agreement achieved in 18 of the 24 exercises.
- The RPD scales need further validation for other types of balance exercises and in individuals with balance disorders.
Introduction
The section discussed the importance of balance and vestibular rehabilitation exercises in improving balance in older adults and those with vestibular disorders. However, there is currently no standard way to measure the intensity of balance exercises. In the clinical settings, physical therapists typically progress the challenge of standing balance exercises by reducing sensory input, changing the base of support, and perturbing the balance system. While quantitative methods such as posturography and accelerometers are considered the gold standard for measuring balance control, they are not widely available in clinics. As a result, clinicians often rely on visual observation, timing the duration of maintaining stance, or self-report of exercise difficulty. However, these methods may lack precision and reliability. The research aimed to develop a rating method for standing balance exercises and validated two rating of perceived difficulty (RPD) scales against quantitative sway measures. The study also estimated the test-retest reliability of the RPD scales. The availability of an RPD scale could complement current assessment tools used by physical therapists and help in customizing prescription and progression of standing balance exercises.
Methods
Participants
The participant group in this study consisted of 31 men and 31 women, totaling 62 participants, ranging from 18 to 85 years old who were healthy and independently engaged in daily activities. The participants were divided into 4 age groups: young, middle aged, old, and very old. Several exclusion criteria were applied to ensure the participants' suitability for the study. These included the inability to stand for 3 minutes without rest, clinically significant somatosensory loss, visual acuity worse than 20/40, peripheral vestibular disorders, a diagnosis of benign paroxysmal positional vertigo, history of neurological or orthopedic disorders, excessive weight, cognitive impairment, history of recurrent falls, pregnancy, or the use of an assistive device for ambulation. The study protocol was approved by the Institutional Review Board of the University of Pittsburgh, and all participants provided informed consent prior to their involvement in the research.
Experimental Procedure
The experimental procedure involved eligible participants who completed Activities-Specific Balance Confidence Scale questionnaires and Functional Gait Assessments to get the baseline for their characteristics. Gait speed was measured over 6m, and the average of three trials was calculated. Participants were tested during two experimental visits, one week apart, where they performed 2 sets of 24 static standing balance exercises. The exercises were randomized and included factors such as surface, vision, base of support, and head movements. Participants stood barefoot and used an AIREX Balance Pad. They also performed different foot positions and wore opaque goggles. Participants moved their head in yaw and pitch directions. Data collection stopped if participants lost their balance based on specific failure criteria, including stepping out of position, changing position of feet or arms, or touching external support for balance. Participants provided ratings of perceived difficulty (RPD) using numerical and qualitative scales after each exercise. The numerical scale ranged from 0-10, with higher scores indicating higher difficulty. The qualitative scale had 5 color-coded levels, with verbal anchors ranging from "completely steady" to "about to fall." The RPD-qualitative scale was transformed into numbers for statistical analysis on a scale of 1-5. Participants were instructed to report a number and a letter indicating the difficulty of maintaining balance. The scales were placed on the side wall for easy viewing, and ratings were provided for each trial attempted. At the end of the experiment, participants were asked which scale they preferred.
Postural Sway Measures
The Postural Sway Measures section of the research paper describes the methodology used to measure trunk sway during exercises. Each participant had an inertial measurement unit (IMU) attached to their lower back at the level of the iliac crest. The IMU recorded trunk angular displacement and velocity in the pitch and roll directions, as well as linear acceleration in the anteroposterior and mediolateral directions. The sampling rate was 100 Hz, and sway measures were recorded for 35 seconds, with the first 5 seconds removed to eliminate initial balance effects. Summary measures of trunk sway were calculated from the remaining 30-second time series, which were low-pass filtered using a Butterworth filter with a cutoff frequency of 3 Hz. Each trial was visually inspected using MATLAB software to ensure no extraneous movements were present. Incomplete trial data were treated as missing data in the analysis. For this study, the Root-Mean-Square (RMS) of the trunk angular displacement and velocity were calculated and used for analysis.
Data Analyses
The data analyses section of the research paper focused on the comparison of participants' demographic characteristics between different groups and the assessment of concurrent validity of RPD scales. For dependent variables that were continuous and normally distributed, a 1-way analysis of variance was used to compare demographic characteristics between groups, and post hoc comparisons were conducted to evaluate pairwise differences among the groups. The Sidak approach was utilized to control for type I error.
In cases where dependent variables were continuous but not normally distributed, the Kruskal-Wallis test was used to compare demographic characteristics between groups, and the Dunn procedure was employed for pairwise comparisons, with a Bonferroni correction for multiple comparisons.
To assess the concurrent validity of the RPD scales, the researchers examined the relationship between each RPD scale and each postural sway variable using a general linear model method. This approach accounted for correlated data within each participant, specifically the scores of RPD and postural sway measures. As there were up to 24 observations per participant (one for each exercise), the regression model treated participant as a fixed effect to remove variation attributed to them. The mean values of RPD and postural sway measures from all four trials were used in the model for each exercise.
Two scales were used: a numerical scale based on the OMNI Perceived Exertion Scale and a qualitative scale based on previous work by Espy et al. The outcome variable was the rating of perceived difficulty (RPD), and participant and postural measures were entered as covariates in the analysis. Correlation coefficients were calculated to examine the relationship between the RPD scales and postural sway measures. Visually examining the data, a curvilinear association was observed, particularly in older age groups. Therefore, both linear and semilogarithmic relationships were tested, and it was found that the logarithmic relationship provided a slightly better fit. The RPD scales were further validated against different balance exercise conditions, such as different surfaces, vision status, base of support, and head movements. Nonparametric tests were used to examine the relationship between the exercise factors and RPD scores. To assess the test-retest agreement (reliability) of the RPD scales, a weighted kappa with linear weights was used. Point estimates and 95% confidence intervals were calculated to determine the reliability of the scales.
Results
The study included 62 participants who completed the study out of the initial 72 people who underwent onsite screening. The 10 participants who did not complete the study had a mean age of 64 years. Eight participants were excluded due to not passing the inclusion criteria, and two participants did not return for the experimental visits. The data showed that the very old group had a slower gait speed compared to the young and middle-aged groups. The young group had a lower body mass index compared to the old and very old groups. The very old group also had lower balance confidence and functional gait assessment scores compared to the other groups.
The study found that the participants had greater RPD scores on a foam surface, with their eyes closed, in a semitandem stance, and with yaw and pitch head movements. There was a significant positive correlation between RPD scores and postural sway measures. When assessing the completion rate of exercises in each group, young and middle aged groups completed the majority of exercises. However, the older aged group had lower completion rate of exercises. The correlation was stronger between RPD scores and roll measurements compared to pitch measurements.
The test-retest agreement of RPD scores was moderate for some exercises within each visit but did not show moderate agreement between visits. The agreement was fair for most exercises within each visit and between visits. The weighted kappa scores for the RPD-qualitative scale ranged from fair to substantial within each visit and between visits, with moderate agreement observed for only a few exercises.
Among the participants, 66.1% preferred the RPD-qualitative scale over the RPD-numerical scale, stating that the statements in the qualitative scale described how they felt.
Discussion
The objective of the study was to validate the RPD-scales with quantitative postural sway measurements. The results showed strong associations between the RPD scales and postural sway measures, indicating that perception of difficulty in performing a balance exercise relates to body sway. The ratings were higher when using foam, reducing visual input, reducing the base of support, and increasing vestibular stimulation. Previous research found a relationship between center of pressure and participants' rating of steadiness, but only four balance exercises were studied.
The study found a logarithmic relationship between perception of balance and postural sway, suggesting nonlinearity in this relationship. There was a higher correlation between the logarithm of postural sway and perceived difficulty compared to untransformed sway values, particularly in older adults. The correlation between Rating of Perceived Difficulty (RPD) scales and postural sway was stronger for roll position and velocity than for pitch position and velocity, potentially due to participants basing their RPD more on their perception of sway in the ML direction. The test-retest reliability of RPD scales was moderate and improved with repeated use, suggesting calibration of ratings. The RPD-qualitative scale had lower agreement values than the RPD-numerical scale, potentially due to the shorter scale having a greater prevalence effect. The weighted kappa values for RPD scales were lower compared to ratings of perceived exertion scales for aerobic and resistance exercises, indicating that perception of standing balance performance may be more complex or variable. The study had limitations, including possible fatigue effects and nonindependence of ratings. The reliability and validity of RPD scales should be investigated in populations that would benefit from balance rehabilitation.
The ease of the use of RPD scales compared to the use of technology for assessment of postural sway in the clinical setting was the main strength of the study. The use of an RPD scale during a routine physical therapy visit can enhance individualization of exercise program that can tailor to the patient’s needs, and the RPD scale can also be safely used in home settings for a feedback to the physical therapist for overall difficulty or intensity of the balance exercise program. Overall, the study provides insights into the relationship between perception of balance, postural sway, and RPD scales, but further refinement and investigation are needed.
Opportunities for Future Research
1. Further research could investigate the validity and reliability of the RPD scales in populations with balance disorders, such as individuals with vestibular disorders or older adults.
2. Future studies could validate the RPD scales for other types of balance exercises, including modified center of gravity exercises, weight shifting exercises, and gait exercises.
3. Additional research can explore the relationship between perceived difficulty and postural sway for different types of balance exercises, using a variety of instrumented measures.
4. Further investigation is needed to determine the optimal granularity and descriptors for the RPD scales to improve their reliability and agreement.
5. Future studies could examine the relationship between perceived difficulty and other outcome measures of balance, such as functional balance tests or measures of falls risk.
6. Research could explore the use of the RPD scales in other rehabilitation settings or populations, such as individuals with neurological disorders, to assess the intensity and progression of balance exercises.
