INTRODUCTION
The progressive nature of Parkinson disease (PD) leads to significant movement-related impairments.1 Gait difficulty, including freezing of gait, and postural instability are common among people with PD and are highly associated with falls.2,3 Falls occur in up to 70% of individuals with PD, with potential sequelae including mobility restrictions, fractures, and mortality.4–7 Injuries related to falls significantly increase medical costs for those with PD.8 Given the devastating effects of falls, it is critical that individuals at risk for falls be accurately identified so that they can receive interventions directed at improving gait and balance and reducing fall risk.
Previous research shows that the best predictor of falls is a prior history of falls,9 but this relies on accurate self-report of past events. It is preferable for clinicians to be able to directly measure a person's performance in order to evaluate fall risk. Recent work suggests that some clinical balance tests such as the Balance Evaluation Systems Test (BESTest) and Mini-BESTest may be useful for identifying those who will fall in the next 6 months.10 However, these and other clinical balance tests are often time consuming and therefore may be impractical to use on a regular basis. As such, investigation of the clinical utility and predictive value of quick and easy-to-administer clinical tests is warranted.
The Four Square Step Test (4SST), first described by Dite and Temple,11 is a quick and simple clinical test that requires an individual to rapidly step over obstacles in the forward, backward, and sideways directions. We hypothesized that the 4SST may be a useful balance screen in people with PD for the following reasons: (a) it demands complex anticipatory postural adjustments for gait initiation, known to be impaired in people with PD12–15; (b) the requirements for stepping backward and laterally may be particularly challenging for people with PD16,17; and (c) the task requires execution of a complex multistep movement sequence, another area of known difficulty in PD.18–20 To our knowledge, no study to date has evaluated 4SST performance in people with PD. However, 4SST performance has been studied in older adult fallers and nonfallers, individuals with stroke, and individuals with a vestibular disorder.11,21,22
The purpose of this study was to evaluate 4SST performance in individuals with PD. We aimed to (1) establish interrater and test-retest reliability of the 4SST in PD, (2) determine the effectiveness of the 4SST in distinguishing between on and off anti-PD medication conditions, fallers and nonfallers, and freezers and nonfreezers, and (3) describe factors related to and predictive of 4SST performance in PD. We hypothesized that the 4SST would possess high interrater and test-retest reliability, correctly distinguish between on and off anti-PD medication conditions, fallers and nonfallers, and freezers and nonfreezers, and be related to balance, bradykinesia, and freezing of gait in people with PD.
METHODS
Participants
Individuals older than 40 years with idiopathic PD who were already participating in 2 separate, ongoing studies were consecutively recruited.23,24 Participants included in the study must have been diagnosed with “definite” PD, as defined by Racette and colleagues.25 Potential participants were screened and excluded if found to have any of the following: (1) a serious medical condition, (2) history or presence of a neurologic condition other than PD, (3) previous surgical management of PD (eg, deep brain stimulation), or (4) a musculoskeletal injury that considerably inhibited movement. All participants provided written informed consent to participate in the protocol as approved by the Human Research Protection Office of Washington University, St. Louis, Missouri.
Fifty-three individuals (58% male) with idiopathic PD were included in the study. Each participant reported fall history over the past 6 months and, based on this, was classified as a faller if he reported more than 1 fall in that period of time. Participants reporting no falls or only 1 fall in the previous 6 months were classified as nonfallers. Participants also completed the Freezing of Gait Questionnaire (FOG-Q), a self-assessment of walking performance during their self-perceived worst state, how gait impairments impact functional independence, whether or not freezing is experienced, and characteristics of freezing episodes.26 A participant was classified as a freezer if he or she gave a rating greater than 1 on item 3 of the FOG-Q, indicating FOG episodes occurring at least once per week. One participant was excluded because of an inability to step in the backward direction secondary to freezing. Demographic information is presented in Table 1. Twenty-eight participants (54% male), a subgroup of the original sample, were chosen to complete the 4SST on and off anti-PD medication to determine whether the 4SST possessed the ability to distinguish between medication conditions.
;)
Table 1: Demographic Information of Participants
Outcome Measures
Four Square Step Test
Performance of this test was carried out as originally described by Dite and Temple.11 Meter sticks, set up to appear like a crosshair, were used as the obstacles over which participants would step (Figure 1). These meter sticks were not fixed to the surface. Participants faced the same direction when stepping to the different squares, and both feet had to touch each square before the next movement. Stepping was completed in the following sequence: forward, right, backward, left, right, forward, left, backward. Timing began when the first foot hit the second square and stopped when the last foot returned to the final square. Successful completion of a trial was a trial in which the meter sticks were not touched during performance of the test. The 4SST has been shown to have high interrater and test-retest reliability among the older adults and individuals with vestibular disorders.11,22
Figure 1: Diagram illustrating the setup utilized for the Four Square Step Test. Four separate meter sticks (light gray) were arranged on the floor as shown. Arrows illustrate directions of stepping forward, lateral, and backward to complete a full circuit first in the clockwise (black, numbers 1–4) and then in the counterclockwise (dark gray, numbers 5–8) direction.
Movement Disorders Society–Unified Parkinson Disease Rating Scale–III
Motor symptom severity was assessed using the Movement Disorders Society–Unified Parkinson Disease Rating Scale–III (MDS-UPDRS-III).27 From this assessment, we also determined the Hoehn & Yahr stage of each participant. The MDS-UPDRS-III was administered by a physical therapist (R.P.D.) who was trained using the official MDS-UPDRS training video.28
Mini-Balance Evaluations Systems Test
The Mini-BESTest is an instrument used to assess balance through evaluation of postural stability during performance of 14 dynamic tasks.29 Investigators have previously demonstrated high interrater and test-retest reliability, when using the Mini-BESTest to evaluate balance in individuals with PD.30
Five Times Sit to Stand Test
Since it is possible that the 4SST and Five Times Sit to Stand (FTSTS) test might be measuring similar constructs, the FTSTS was included in the assessment battery. The FTSTS test measures how quickly one can go from sitting to standing 5 consecutive times without using the upper extremities. Balance and bradykinesia are predictive of FTSTS performance in PD.31 This test has also been shown to have high interrater and test-retest reliability in PD.31
Six-Minute Walk Test
Endurance was measured using the Six-Minute Walk Test (6MWT), which has been shown to be highly reliable when assessing people with PD.32 Participants were asked to cover as much ground as they could in 6 minutes while walking, and if necessary, assistive devices were used. Distance covered was measured to the nearest meter.
Nine Hole Peg Test
The Nine Hole Peg Test (9HPT) was used as a measure of bradykinesia.33,34 Participants were instructed to place 9 pegs into 9 holes, 1 at a time, as quickly as possible. The nondominant hand was used to stabilize the peg board if necessary. Two trials were completed using the dominant hand, and the mean of these 2 trials was calculated. This test is highly reliable when examining upper extremity function in people with PD.33
Procedures
All outcome measures were assessed while participants were on anti-PD medication. In addition, a subset (n = 28) of this group who were those recruited for one of the 2 studies mentioned previously was also tested while off anti-PD medication. Participants were considered off-medication if the last medication dose was administered greater than 12 hours prior to evaluation. A physical therapist instructed participants in correct performance of the 4SST. Following instruction, 1 untimed practice trial and 2 timed trials of the 4SST were completed. Two raters timed the trials. Interrater reliability was determined comparing the times recorded by the 2 raters for the first trial performed, while test-retest reliability was determined comparing the times of the first and second trials collected by the primary rater. Because it was our aim to determine whether 4SST performance could be distinguished between off-and on-medication states, only 4SST times were obtained in the off-medication state. After completing off-medication testing, participants were allowed to take their normal dose of anti-PD medication. For participants on-medication, we administered the measures in the following order: MDS-UPDRS-III, 4SST, Mini-BESTest, FTSTS test, 9HPT, FOG-Q, and 6MWT. All outcome measures were administered by the same rater.
Data Analysis
Descriptive statistics were calculated using only the first trial of the 4SST because this was deemed most clinically relevant by the authors. Intraclass correlation coefficients (ICC) were used to describe interrater (ICC 1,1) and test-retest (ICC 2,1) reliability of the 4SST. For all analyses beyond reliability and the off/on-medication comparison, only the first timed trial of the 4SST performed on-medication was used. Since 4SST times were not normally distributed, we used nonparametric statistics to examine differences between conditions or groups. To determine differences between off-and on-medication 4SST times, we used a Wilcoxon signed rank test. To determine whether there were differences in 4SST performance between fallers and nonfallers and between freezers and nonfreezers, we used Mann-Whitney U tests. Spearman correlation coefficients were used to describe relationships between the 4SST and all other outcome measures. Statistical significance for analyses was set at α ≤ .05. Outcome measures that were significantly correlated with the 4SST were entered into an exploratory, simultaneous regression model to determine factors predictive of 4SST performance. In cases of multicolinearity, the variable with the highest correlation with the 4SST was retained and the collinear variable was removed from the analysis. In the final model, only those factors that significantly contributed (P < 0.05) were retained. Finally, a cutoff score for 4SST performance in fallers versus nonfallers was determined using receiver operating characteristic (ROC) curves. This cutoff score was chosen on the basis of the minimum value of (1 − sensitivity)2 + (1 − specificity)2.35 From this cutoff score, positive (LR+) and negative (LR−) likelihood ratios and pre- and posttest probabilities were calculated. The ROC curve of the 4SST was compared with that of the Mini-BESTest for this sample. Data analysis was conducted using NCSS software version 7.1.19 (NCSS Software, Kaysville, Utah).
RESULTS
Interrater reliability, tested only on-medication, was high (ICC = 0.99). Test-retest reliability was high when testing participants both off-(ICC = 0.90) and on-(ICC = 0.78) medication.
The median 4SST performance time on-medication was 9.52 (95% Confidence limits: 9.03–10.54) seconds (Table 2). Figure 2 shows the frequency distribution for 4SST scores on-medication for the full sample, with 4SST times binned into 1-second intervals. The 4SST performance was significantly different off-medication compared with on-medication (P = 0.03). There were no significant differences in 4SST performance between freezers and nonfreezers (P = 0.08) and between fallers and nonfallers (P = 0.06) (Table 2). All outcome measures were significantly related to 4SST performance on-medication (Table 3), most notably the Mini-BESTest, MDS-UPDRS-III, age, 9HPT, and FTSTS.
Figure 2: Frequency histogram showing distribution of Four Square Step Test (4SST) times for the full sample when tested on-medication. Times are grouped into 1-second bins.
Table 2: 4SST Performance Times (Group Comparisons)
Table 3: Spearman Correlations of Outcome Measures With 4SST
Age, MDS-UPDRS-III, Mini-BESTest, FTSTS, 9HPT, FOG-Q, and 6MWT were all entered into a simultaneous regression model, as all of these factors were significantly correlated with 4SST performance. However, because of issues with multicolinearity with the Mini-BESTest, age and MDS-UPDRS-III were removed from the model. Of the remaining 5 variables entered, only 3 contributed significantly to the model. Thus, the final model included only the Mini-BESTest, 9HPT, and FTSTS as shown in Table 4. This model explained 56% of the variance in 4SST performance.
Table 4: Final Regression Model for 4SST
The ROC curves of the 4SST and the Mini-BESTest for discriminating between fallers and nonfallers are displayed in Figure 3. The AUC for the 4SST was 0.65 (95% CI: 0.43–0.80) compared with an AUC of 0.80 for the Mini-BESTest. Maximizing sensitivity and specificity, the cutoff score for the 4SST was determined to be 9.68 seconds (sensitivity = 0.73, specificity = 0.57). Forty-seven percent of our sample scored greater than this cutoff. The positive and negative likelihood ratios for those scoring greater than the cutoff were 1.7 (95% CI: 1.00–2.73) and 0.48 (95% CI: 0.17–1.27), respectively. The posttest probability of a fall for those with scores greater than the cutoff was 31%, compared with a 21% pretest probability.
Figure 3: Receiver operating characteristic curves of the Four Square Step Test (4SST) and Mini-BESTest.
DISCUSSION
To our knowledge, this is the first study to describe 4SST performance in people with PD. According to our results, there is high interrater and test-retest reliability for the 4SST when testing individuals with PD on and off anti-PD medication. Another finding was that people with PD perform the 4SST more slowly off-medication than they do when on-medication; however, there was no difference in 4SST performance for freezers/nonfreezers and fallers/nonfallers. Finally, all measures employed were significantly correlated with 4SST performance. Measures reflective of balance and bradykinesia combined to explain 56% of the variance in 4SST performance in people with PD.
Interrater reliability (ICC = 0.99), when testing individuals with PD on anti-PD medication, matches that noted by Dite and Temple11 in the original description of the 4SST. The test-retest reliability when testing people on-medication (ICC = 0.78) was lower than expected. While the designs for assessing test-retest reliability were similar, Whitney and colleagues22 reported much higher reliability (ICC = 0.93) when testing individuals with a vestibular disorder. In our study, it is possible that a practice effect could have led to consecutive improvements in 4SST performance times from the practice trial and then to the first and second trials, lowering agreement between 4SST times. Because we were limited in our ability to have the participants return for a follow-up visit, we could not capture test-retest reliability with more than 1 or 2 minutes between trials.
On-medication, the median time taken by our full sample to complete the 4SST was 9.52 seconds, which is faster than that of healthy older individuals who had at least 1 fall in the previous 6 months, individuals poststroke, and individuals with a vestibular disorder.11,21,22 It is particularly interesting that on average our group with PD was faster than healthy older individuals who had at least 1 fall in the previous 6 months. This difference could be related to the fact that the sample of healthy elders with a history of at least 1 fall was heterogeneous because of inclusion of individuals with varied histories such as stroke, cardiovascular disorders, and neurologic disorders.11 The 4SST performance in PD may also be enhanced by the visual cues provided by the apparatus used for the test. Investigators have shown that visual cues improve gait initiation in the forward direction.36 We speculate that these cues may also translate to improved stepping in the backward and lateral directions and as such the grid used for the 4SST may provide visual cues that enhance performance of stepping during the test. However, it is important to note that Dite and Temple11 reported that a group of 27 healthy older adults, a homogenous sample without comorbidities or history of falls, had a mean 4SST time of 8.70 seconds. This suggests that while the meter sticks may have improved performance for those with PD, there is still evidence of impairment relative to healthy controls.
Foreman and colleagues37(p169) stated that there is a need for clinical balance tests that are “responsive to changes in performance within individuals (medication status) and responsive to differences between individuals (e.g. fallers versus non-fallers).” We studied the responsiveness of the 4SST to changes in medication status in those with PD, in addition to the responsiveness to differences between freezers and nonfreezers and between fallers and nonfallers with PD. The results of this study also indicate that people with PD demonstrate different 4SST performance when off anti-PD medication compared to on. We found improved 4SST performance while on-medication, a finding in accordance with Morris and colleagues, who found improved Timed Up and Go performance on-medication versus off.38 In our sample, the 4SST did not distinguish between fallers and nonfallers and between freezers and nonfreezers. This may be related to our relatively small sample sizes of fallers and freezers, which reduces our power to detect differences.
Our results demonstrated that all measures employed in this study were significantly related to 4SST performance in people with PD. Most notably, the Mini-BESTest score demonstrated a significant moderate correlation to 4SST performance. This is not surprising as good balance is necessary to complete the multidirectional stepping test in a timely fashion. The other correlation worth noting is the significant relationship between the 9HPT and 4SST. Earhart and colleagues33 noted that bradykinesia was a significant predictor of 9HPT performance time in people with PD. Because of the relationship between the 9HPT and 4SST, we suspect that bradykinesia might also be influencing 4SST performance in this population. We also noted that the Mini-BESTest, 9HPT, and FTSTS combined to explain 56% of the variance in 4SST performance. This is very similar to results reported by Duncan et al,31 who studied factors predictive of FTSTS performance in people with PD. Because both the 4SST and FTSTS are timed tests of mobility, the results noted by Duncan et al31 and those reported in this study suggest that balance and bradykinesia play significant roles in timed tests of mobility in people with PD.
Regarding the detection of fallers, the 4SST demonstrated moderate levels of sensitivity (0.78) and specificity (0.57) with a cutoff time of 9.68 seconds. A 4SST time above the cutoff does not immediately suggest that the person with PD is at significant risk for falls. This notion is supported by the fact that when an individual with PD completes the 4SST in more than 9.68 seconds (LR+ = 1.7), the posttest probability of being a faller is 31% compared with a pretest probability of 21%. A 4SST completion time less than the cutoff (LR− = 0.48) yields a posttest probability of 11%. Based on our analysis, it is evident that the Mini-BESTest outperforms the 4SST in identifying fallers and nonfallers with PD. Leddy and colleagues30 demonstrated that the Mini-BESTest (sensitivity = 0.88, specificity = 0.78) and BESTest (sensitivity = 0.84, specificity = 0.76) had better predictive ability than the 4SST in the present study. These findings were echoed in a prospective study of fall risk in PD in which the Mini-BESTest (sensitivity = 0.86, specificity = 0.78) and BESTest (sensitivity = 0.93, specificity = 0.84) outperformed other balance measures when identifying fallers 6 months after assessment.10 While the 4SST is similar in its predictive abilities to other quick tests of motor function, the 4SST should not replace more extensive balance testing when attempting to determine fall risk in people with PD.31,37
The results of this study should be interpreted in light of several limitations. First, this pilot study included a small sample size of people with mild to moderate PD among which there were small numbers of fallers and freezers. The results demonstrating that the fallers and freezers performed the 4SST more slowly than nonfallers and nonfreezers, respectively, should be interpreted with caution. Second, all analyses other than test-retest reliability utilized only 1 timed trial of the 4SST. This was done to reflect the clinical practice of rehabilitation professionals where it is likely that only 1 trial of the 4SST would be collected. However, somewhat-different results could have been obtained if the results of several trials had been averaged. Finally, we measured test-retest reliability with only minutes between 4SST trials and recommend that investigators measure test-retest reliability of the 4SST in people with PD with perhaps days or a week between administrations of the test. In the future, investigators should study responsiveness of the 4SST to change over time, or after intervention, and may also consider examining the combined ability of the 4SST and other quick tests of motor function in predicting who will fall with PD.
CONCLUSION
The 4SST, when used in people with PD, is able to distinguish between off and on anti-PD medication states, but not between freezers and nonfreezers and between fallers and nonfallers. Mini-BESTest, 9HPT, and FTSTS scores were most predictive of 4SST performance. Because of its limited ability in accurately predicting falls in those with PD, we do not recommend the use of the 4SST in lieu of other balance measures such as the Mini-BESTest.
ACKNOWLEDGMENT
The authors thank Marie McNeely, PhD, for her assistance with timing of the 4SST.
REFERENCES
1. Speelman AD, van de Warrenburg BP, van Nimwegen M, Petzinger GM, Munneke M, Bloem BR. How might physical activity benefit patients with
Parkinson disease? Nat Rev Neurol. 2011;7(9):528–534.
2. Bloem BR, Grimbergen YA, Cramer M, Willemsen M, Zwinderman AH. Prospective assessment of falls in Parkinson's disease. J Neurol. 2001;248(11):950–958.
3. Bloem BR, Hausdorff JM, Visser JE, Giladi N. Falls and freezing of gait in Parkinson's disease: a review of two interconnected, episodic phenomena. Mov Disord. 2004;19(8):871–884.
4. Bloem BR, Steijns JA, Smits-Engelsman BC. An update on falls. Curr Opin Neurol. 2003;16(1):15–26.
5. Snijders AH, Nonnekes J, Bloem BR. Recent advances in the assessment and treatment of falls in Parkinson's disease. F1000 Med Rep. 2010;2:76.
6. Sato Y, Kaji M, Tsuru T, Oizumi K. Risk factors for hip fracture among elderly patients with Parkinson's disease. J Neurol Sci. 2001;182(2):89–93.
7. Wenning GK, Ebersbach G, Verny M, et al. Progression of falls in postmortem-confirmed parkinsonian disorders. Mov Disord. 1999;14(6):947–950.
8. Pressley JC, Louis ED, Tang MX, et al. The impact of comorbid disease and injuries on resource use and expenditures in parkinsonism. Neurology. 2003;60(1):87–93.
9. Pickering RM, Grimbergen YA, Rigney U, et al. A meta-analysis of six prospective studies of falling in Parkinson's disease. Mov Disord. 2007;22(13):1892–1900.
10. Duncan RP, Leddy AL, Cavanaugh JT, et al. Accuracy of fall prediction in
Parkinson disease: six-month and 12-month prospective analyses. Parkinsons Dis. 2012;2012:237673.
11. Dite W, Temple VA. A clinical test of stepping and change of direction to identify multiple falling older adults. Arch Phys Med Rehabil. 2002;83(11):1566–1571.
12. Burleigh-Jacobs A, Horak FB, Nutt JG, Obeso JA. Step initiation in Parkinson's disease: influence of levodopa and external sensory triggers. Mov Disord. 1997;12(2):206–215.
13. Hass CJ, Waddell DE, Fleming RP, Juncos JL, Gregor RJ. Gait initiation and dynamic
balance control in Parkinson's disease. Arch Phys Med Rehabil. 2005;86(11):2172–2176.
14. Rocchi L, Chiari L, Mancini M, Carlson-Kuhta P, Gross A, Horak FB. Step initiation in Parkinson's disease: influence of initial stance conditions. Neurosci Lett. 2006;406(1/2):128–132.
15. Rosin R, Topka H, Dichgans J. Gait initiation in Parkinson's disease. Mov Disord. 1997;12(5):682–690.
16. Gantchev N, Viallet F, Aurenty R, Massion J. Forward versus backward oriented stepping movements in Parkinsonian patients. Motor Control. 2000;4(4):453–468.
17. King LA, Horak FB. Lateral stepping for postural correction in Parkinson's disease. Arch Phys Med Rehabil. 2008;89(3):492–499.
18. Benecke R, Rothwell JC, Dick JP, Day BL, Marsden CD. Disturbance of sequential movements in patients with Parkinson's disease. Brain. 1987;110(Pt 2):361–379.
19. Weiss P, Stelmach GE, Hefter H. Programming of a movement sequence in Parkinson's disease. Brain. 1997;120(Pt 1):91–102.
20. Yaguez L, Lange HW, Homberg V. Differential effect of Huntington's and Parkinson's diseases in programming motor sequences of varied lengths. J Neurol. 2006;253(2):186–193.
21. Blennerhassett JM, Jayalath VM. The
Four Square Step Test is a feasible and valid clinical test of dynamic standing
balance for use in ambulant people poststroke. Arch Phys Med Rehabil. 2008;89(11):2156–2161.
22. Whitney SL, Marchetti GF, Morris LO, Sparto PJ. The reliability and validity of the
Four Square Step Test for people with
balance deficits secondary to a vestibular disorder. Arch Phys Med Rehabil. 2007;88(1):99–104.
23. Dibble LE, Cavanaugh JT, Earhart GM, Ellis TD, Ford MP, Foreman KB. Charting the progression of disability in
Parkinson disease: study protocol for a prospective longitudinal cohort study. BMC Neurol. 2010;10:110.
24. Duncan RP, Earhart GM. Randomized controlled trial of community-based dancing to modify disease progression in
Parkinson disease. Neurorehabil Neural Repair. 2012;26(2):132–143.
25. Racette BA, Rundle M, Parsian A, Perlmutter JS. Evaluation of a screening questionnaire for genetic studies of Parkinson's disease. Am J Med Genet. 1999;88(5):539–543.
26. Giladi N, Tal J, Azulay T, et al. Validation of the freezing of Gait Questionnaire in patients with Parkinson's disease. Mov Disord. 2009;24(5):655–661.
27. Goetz CG, Tilley BC, Shaftman SR, et al. Movement Disorder Society–sponsored revision of the Unified Parkinson's Disease Rating Scale (MDS-UPDRS): scale presentation and clinimetric testing results. Mov Disord. 2008;23(15):2129–2170.
28. Goetz CG, Stebbins GT, Chmura S, Fahn S, Klawans H, Marsden CD. Unified Parkinson's Disease Rating Scale (UPDRS) Training Program.
http://www.movementdisorders.org/updrs/. Published June 11, 2012. Accessed August 13, 2012.
29. 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):323–331.
30. 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):90–97.
31. Duncan RP, Leddy AL, Earhart GM. Five times sit-to-stand test performance in Parkinson's disease. Arch Phys Med Rehabil. 2011;92(9):1431–1436.
32. Steffen T, Seney M. Test-retest reliability and minimal detectable change on
balance and ambulation tests, the 36-Item Short-Form Health Survey, and the Unified
Parkinson Disease Rating Scale in people with parkinsonism. Phys Ther. 2008;88(6):733–746.
33. Earhart GM, Cavanaugh JT, Ellis T, Ford MP, Foreman KB, Dibble L. The 9-hole PEG test of upper extremity function: average values, test-retest reliability, and factors contributing to performance in people with
Parkinson disease. J Neurol Phys Ther. 2011;35(4):157–163.
34. Oxford Grice K, Vogel KA, Le V, Mitchell A, Muniz S, Vollmer MA. Adult norms for a commercially available Nine Hole Peg Test for finger dexterity. Am J Occup Ther. 2003;57(5):570–573.
35. Perkins NJ, Schisterman EF. The inconsistency of “optimal” cutpoints obtained using two criteria based on the receiver operating characteristic curve. Am J Epidemiol. 2006;163(7):670–675.
36. Jiang Y, Norman KE. Effects of visual and auditory cues on gait initiation in people with Parkinson's disease. Clin Rehabil. 2006;20(1):36–45.
37. Foreman KB, Addison O, Kim HS, Dibble LE. Testing
balance and
fall risk in persons with
Parkinson disease, an argument for ecologically valid testing. Parkinsonism Relat Disord. 2011;17(3):166–171.
38. Morris S, Morris ME, Iansek R. Reliability of measurements obtained with the Timed “Up & Go” Test in people with
Parkinson disease. Phys Ther. 2001;81(2):810–818.
