Increasing evidence has emerged to support the role of physical exercise for people with Parkinson disease (PwPD),1–4 and current recommended practice is to combine physical exercise and dopaminergic medication for the treatment of motor symptoms.5 Nevertheless, Parkinson disease (PD) is a heterogeneous disorder presenting with varied motor phenotypes as well as a variety of nonmotor signs and symptoms.6 Not only is the disease of a progressive nature, where the severity of symptoms ranges from very mild to severe7 but also various subtypes of PD have been suggested to occur at the early disease stages.8 Therefore, it is unlikely that a given exercise modality benefit all PwPD. This makes it essential to identify disease-specific and personal factors, as well as levels of function, which may affect the outcomes following an intervention.
Exercise interventions range from being delivered in participants' homes or as group training at facilities, with varying degrees of supervision and dosage.9 Moreover, exercise characteristics vary from basic exercises to more complex, specific, and attention-demanding exercises, such as dual tasks.10–12 Although it has been proposed that gait and balance training needs to be challenging to be effective, there is currently no consensus on the optimal ingredients of exercise interventions in PwPD.2
A variety of factors may be important for the response to exercise interventions designed for PwPD. Demographic factors such as older age and/or a history of recurrent falls may have a negative impact on motor learning.13 Similarly, PD-related impairments may restrict the response to exercise.14 People with Parkinson disease who exhibit freezing of gait have been shown to have less effective motor-learning capacities,14 and levodopa medication has been suggested to impair motor learning due to a potential overstimulation of certain regions of the basal ganglia.15 Physical abilities may also play a role. People with Parkinson disease with lower physical functioning may have more room to improve following an intervention, whereas too low physical capabilities may be a limiting factor since motor learning takes a longer time in this population.16
Another area that has received increasing attention with regard to motor learning is that of cognitive abilities. It has been suggested that working memory is essential for sensorimotor adaptation and motor-sequence learning,17 which may be particularly relevant for PwPD where impaired cognition occurs in up to 50% of the population within the first years with the disease.18 Indeed, impaired dual-task abilities, considered a measure of executive functions, are common in this population and have also been linked to increased freezing of gait.19 Finally, factors related to perceived health may also contribute to exercise response in PwPD. Pain, for example, is related to fear avoidance and inactivity,20 whereas fear of falling21 and depression22 have been found to be barriers to exercise in PwPD. Similarly, since higher self-efficacy is related to exercise behavior in this population,23 it is possible that related factors may contribute to exercise response in PwPD.
Taken together, despite the importance of investigating factors that may be associated with responsiveness to exercise interventions in PwPD, there is a lack of studies related to this issue. One reason for this may be that the large sample sizes required for statistically optimal predictor analyses are rarely met in exercise interventions. Nevertheless, considering the gap of knowledge in this field, even studies based on smaller samples may provide important indications, both for the generation of hypotheses for future research and for clinical practice.
We have previously conducted a highly challenging gait and balance training intervention that incorporated dual tasking among PwPD, the HiBalance program.24 The intervention was delivered as a facility-based group training, entailing supervised 1-hour sessions, 3 times per week for 10 weeks, and the main findings were improvements in both gait and balance performance. With this preplanned secondary analysis, we therefore aimed to investigate which factors that best predicted responsiveness of 2 important clinical outcomes belonging to different constructs of mobility in PD25 (gait speed and balance performance, respectively) following participation in the HiBalance program.
The research question was as follows: Which factors related to individual demographics, PD, physical and cognitive abilities, and perceived health best predicted responsiveness of balance performance and gait speed among PwPD following a highly challenging and attention-demanding gait and balance training intervention?
This study was a preplanned secondary analysis of a 2-armed randomized trial for PwPD (Trial registration: NCT01417598), conducted as an exploratory regression analysis investigating factors associated with exercise response in PwPD. The protocol and training program for this trial have been detailed elsewhere.24 , 26 Data were collected between January 2012 and May 2013. The Regional Ethical Board in Stockholm approved this study. All participants gave written informed consent before data collection began.
People with Parkinson disease were recruited via newspaper advertisements and outpatient clinics. Inclusion criteria were a clinical diagnosis of idiopathic PD according to the Queens Square Brain Bank criteria;27 Hoehn and Yahr stage 2 or 3;28 60 years of age or older; the ability to independently ambulate indoors without a walking aid, and at least 3 weeks of stable dopaminergic medication. To increase the ecological validity, only PwPD exhibiting gait and/or balance impairments were included (ie, individuals who would have been considered for rehabilitation in clinical practice by the assessors). Exclusion criteria consisted of cognitive impairments defined by a Mini-Mental State Examination score of less than 2429 , 30 and other medical conditions that might substantially influence gait and balance performance.
The data collection consisted of a structured interview and the assessment of disease severity and physical ability performance in a movement laboratory. All assessments were conducted during the participants ON-state of medication at the same time of day, both at baseline and at the 10-week follow-up. All assessments were conducted by trained physical therapists with clinical experience in neurological rehabilitation, and the participants underwent a practice session before undertaking the gait and balance assessments, which followed a randomized scheme. The participants were asked not to change their dose of medication during the intervention.
Following data collection at baseline, block randomization was used to allocate participants to the experimental group or to the control group (usual care). Both testers and participants were blinded to group allocation during the baseline assessments but not during the follow-up assessments, as some assessors also served as trainers. Only the PwPD who were randomized to the experimental group and completed both the baseline and the 10-week follow-up assessments were included in this secondary analysis.
The experimental group undertook the Hibalance program,26 consisting of 10 weeks (three 60-minute sessions per week) of highly challenging gait and balance training that was conducted in groups of 4 to 7 participants and incorporated dual tasking. The training was performed at 2 sites (north and south) of a hospital by physical therapists with experience in neurologic rehabilitation. The training concept relied upon the continuous progression and adaptation of exercises with regard to the participants' abilities. The progression/adaptation scheme was primarily developed during the planning of the training that occurred between the training sessions; however, minor adjustments were performed during the sessions to optimize the challenge level for each participant. All trainers had been educated in detail about the program's underlying theories and its practical applications during a 2-day workshop.
We defined highly challenging gait and balance training as “exercises inducing intermittent reactive postural adjustments,” for example, during walking on a narrow balance foam, following a sudden stop/turn, or while catching a balance ball. The program consists of 4 training components specific to gait and balance impairments in PwPD: (1) Sensory integration, (2) Anticipatory postural adjustments, (3) Motor agility, and (4) Stability limits.
The training period was divided into 3 blocks to gradually increase the challenge level and movement complexity. During the first block (weeks 1-2), the participants were introduced to the exercises of each training component separately. During the second block (weeks 3-6), cognitively challenging dual-task exercises were gradually introduced and the difficulty level for each training component was increased. In the final block C (weeks 7-10), the difficulty level and the variation were further increased by using exercises combining the different training components, for example, by making sharp turns on an unstable surface.
The dependent variables were percent change in balance performance and gait speed (ie, follow-up − baseline/baseline).25 Balance performance was assessed with the Mini-BESTest,31 which has been found reliable and valid for PwPD with mild to moderate motor severity.32 , 33 This test consists of 14 items, where each item is scored between 0 and 2. The maximal score of the Mini-BESTest is 28 points, and a higher score indicates better balance performance. A 9-meter (active length: 8.3 m) instrumented walkway system (GAITRite, CIR systems, Inc, Franklin, New Jersey) was used for the assessment of gait speed. The participants walked at their comfortable speed; a mean of 6 trials was used for analysis. To avoid effects of acceleration and deceleration, each walk started 3 m in front of the walkway and stopped 3 m behind it.34
Independent Variables (Predictors)
Age, sex, and fall status (recurrent or nonrecurrent fallers) were used as demographic factors. The participants were considered recurrent fallers if they had experienced at least 2 falls in the previous 12 months.
Time since diagnosis, freezing of gait, levodopa-equivalent dosage,35 activities of daily living, and severity of motor symptoms were used as PD-specific factors. Freezing of gait was assessed with item 14 on the original Unified Parkinson's Disease Rating Scale (UPDRS),36 part II, where PwPD were considered freezers if they reported that they experienced any kind of freezing of gait, that is, a score of 1 or greater indicated the occurrence of freezing of gait. Activities of daily living were assessed with the total score of the (UPDRS) part II (maximal score = 52, a higher score signifies more problems). The severity of motor impairments was assessed with the (UPDRS) part III, which contains 27 items that cover the domains of bradykinesia, tremor, rigidity, and postural instability (maximal score = 108, a higher score reflects more severe symptoms).
The time to complete the Timed Up and Go (TUG) test and the dual-task gait speed (measured with the instrumented walkway system as described previously) were used as outcomes for physical ability. The TUG is a test of physical mobility and measures the time it takes to stand up from a standard arm chair, walk 3 m, turn 180°, walk back, and sit down again. During the dual-task gait-speed testing, the participants simultaneously performed a cognitive task while walking over the instrumented mat. The cognitive task entailed the reciting of alternate letters of the alphabet,37 and the accuracy of this task was quantified as the mean percent error (number of errors/number of letters recited). The participants were instructed to pay equal attention to both tasks at all times. We considered it clinically meaningful to investigate if a clinical outcome belonging to one construct may provide information regarding the exercise response of a related construct. Therefore, baseline gait speed was used as a factor of improved balance performance, whereas the baseline Mini-BESTest score was used as a factor of improved gait speed.
To gain more insight into cognitive abilities, we evaluated the performance of the cognitive task as a single task (ie, 3 trials while seated), the absolute cognitive dual-task interference (ie, the difference in accuracy of the cognitive task while seated and walking), and absolute dual-task interference while walking (ie, absolute difference in walking speed between walking at a comfortable speed and while walking and performing the cognitive task). These outcomes reflect the interference of walking on the cognitive task, and the interference of the cognitive task on gait performance. For each trial entailing the cognitive task (both during single-task and dual-task conditions), the participants received different starting letters following a structured scheme.
All of the subcategories of the 36-Item Short Form Health Survey38 (SF-36) as well as concerns about falling (assessed with the Falls-Efficacy Scale International)39 were used as factors for perceived health. The SF-36 is a self-assessed questionnaire consisting of 8 subcategories: (1) limitations in physical activities because of health problems, (2) limitations in usual role activities because of emotional problems, (3) limitations in usual role activities because of physical health problems, (4) bodily pain, (5) general health perceptions, (6) vitality (energy and fatigue), (7) limitations in social situations because of physical or emotional problems, (8) general mental health (psychological distress and well-being). Each subcategory is rated from 0 to 100, where a higher score signifies less impairment. The Falls-Efficacy Scale International is a self-assessed questionnaire entailing 16 items covering a variety of daily life situations. Each item is rated from 1 to 4, and the results are added up to a total score ranging from 16 to 64 (lower = better).
The STATISTICA software (Statsoft, version 13, Tulsa, Oklahoma) was used for the statistical analyses. Multiple linear regressions were performed to investigate the association between the independent variables and responsiveness to balance training. Initially, univariate analyses (simple unadjusted linear regression models) were performed to study the effect of each independent variable on the respective outcome. The cutoff to include factors in the multivariate analyses was set to P ≤ 0.2. Because of the small sample size, the α level was set to P = 0.10 for the parameters in the final model in order to reduce the risk of type II error.40 To avoid collinearity between the factors, correlations were conducted and factors with a correlation of greater than 0.6 were considered to be collinear, whereby only the factor with the strongest P value was included in the final model. Thereafter, the manual backward elimination of factors was conducted until all the individual parameters in the model had a P value smaller than 0.10. As the sample size was small in relation to the number of included variables, only the adjusted R 2 (which has been recommended for small sample sizes in linear regression analyses) was used to explain the variance of the dependent outcomes.41
Fifty-one PwPD were included in the experimental group and 4 people dropped out of the program (Figure 1). The PwPD who completed the training program and were measured at baseline and the 10-week follow-up were included in the analyses (n = 47). Participant characteristics can be found in Table 1. Forty-one (87%) of the participants improved their balance performance, and for the whole group (n = 47), there was a mean improvement of 17% (range: −6% to 83%) of the Mini-BESTest score following participation in the HiBalance program. Thirty-five (76%) of the participants improved their gait speed and for the whole group (n = 46), there was a mean improvement of 9% following the training (range: −24% to 56%).
Factors of Improved Balance Performance
Following the univariate analyses, 11 factors met the inclusion criteria (Table 2). Three of these factors were ruled out because of collinearity with other included factors, resulting in 8 remaining factors that were included in the initial multiple linear regression model. Through a manual backward elimination procedure, the following factors were removed: dual-task gait speed, limitations in social situations because of physical or emotional problems, freezing of gait, body pain, and limitations in physical activities because of health problems.
The final model contained 3 factors that were independently associated with change in balance performance at follow-up (Table 3). These factors explained 35% (adjusted R 2 = 0.35, R 2 = 0.39) of the variance. The strongest independent factor was a low rating of self-perceived general health, followed by a longer time to complete the TUG test and a greater amount of errors on the cognitive task.
Factors for Improved Gait Speed
A total of 12 factors met the inclusion criteria (Table 4). Four of these factors were ruled out because of collinearity with other included factors. An initial multiple linear regression analysis was conducted which, following the manual backward elimination procedure, resulted in a final model containing 4 factors that were independently related to improved gait speed following participation in the HiBalance program. These factors explained 49% (adjusted R 2 = 0.49, R 2 = 0.51) of the variance of change in gait speed (Table 5). Slower dual-task gait speed was the strongest independent factor, followed by greater dual-task interference, a longer time to complete the TUG test, and greater amount of errors on the cognitive task.
We aimed to investigate which factors best predicted responsiveness of 2 important clinical outcomes—balance performance and gait speed, which represent different constructs of mobility in PD.25 The strongest predictors of improved balance performance were related to physical ability, perceived health, and cognitive ability, whereas the strongest factors for predicting improved gait speed were related to cognitive ability.
Despite certain overlap between the 2 models, the results highlight that while both balance performance and gait speed are related to physical ability, different factors may be particularly important for the improvement of either outcome. Nevertheless, the TUG test and the performance of the cognitive task were independently associated with improvements of both balance performance and gait speed. The finding that a longer time to complete the TUG test at baseline was related to improvements following training was not surprising as PwPD with worse mobility might have more room for improvement. However, the finding that people with worse, rather than better, cognitive functioning responded better to the training was surprising as better cognitive functioning has been related to higher rate of motor learning.17 In particular, this result should be viewed in relation to the elements of the training program, which was of a challenging character and incorporated a combination of single- and dual-task exercises. Indeed, PwPD with mild cognitive impairment have previously shown impaired gait in comparison with cognitively intact PwPD, differences that were increased during dual-task gait.42 Although this may indicate that the PwPD with worse cognitive functioning in the current study had more room to improve with regard to gait speed, the results also suggest that the dual-task exercises did not interfere with motor learning. Taken together, the findings that the PwPD who performed worst on measures related to cognitive and physical ability responded best to the training suggest that challenging rehabilitation, including dual-task exercises, can be delivered to this population at a high dosage with beneficial results.
Previous research has found that high exercise self-efficacy, education, and older age but not impairment were associated with regular exercise behavior among PwPD.23 However, to our knowledge, studies investigating which factors best predict rehabilitation outcomes are scarce in general and nonexistent among PwPD. Nevertheless, the subgroup analysis from a large randomized trial designed to reduce falls among PwPD found results that may complement our findings.43 That study found a significant falls reduction in PwPD with mild disease severity compared with those with a more progressed severity. However, that study used a minimally supervised home exercise intervention, where only 13% of the training sessions were supervised. Conversely, in the HiBalance program, there were always 2 physical therapists present during the training sessions to ensure a progressive and highly challenging training, as well as safety. Although the 2 studies differ with regard to the construct of the outcomes, as well as exercise modality, these findings may indicate that PwPD with more severe disabilities need to be supervised while performing exercise in order to benefit, whereas PwPD with less severe impairments may be better suited to adequately perform exercises with less supervision. Nevertheless, findings from a recent meta-analysis suggest that supervised training at facilities is preferable for long-term gains in gait and balance abilities.4 More research is needed in this field to enable robust conclusions regarding specific exercise modalities for different subgroups of PwPD.
This study had several limitations. First, due to logistic challenges, the assessors were not blinded to group allocation at follow-up, which may have affected the outcomes. Second, as the study was of an exploratory nature and with a small sample size in relation to the number of included variables, the results need to be interpreted with caution as the model may be susceptible to being overfit. However, for linear regressions, it has recently been suggested that as few as 2 subjects per variable is acceptable with an estimated bias of less than 10% when the adjusted R 2 is used.41 Third, as this was a complex and highly challenging intervention, focusing specifically on gait and balance training incorporating dual-task exercises, the results can be generalized only to this exercise modality. Finally, the intervention was delivered at a high dosage and conducted within the strict frames of a randomized trial; therefore, the results can be generalized only to PwPD with mild to moderate severity without substantial cognitive impairment (Mini-Mental State Examination score of >24) or severe freezing of gait. Nevertheless, this was the first study investigating factors potentially important for response to exercise in PwPD, meaning that initial insights concerning this area are provided. Moreover, the findings that poor performance of factors related to cognitive ability was associated with improved gait speed and balance performance suggest that future exercise interventions designed for PwPD may consider including people with cognitive deficits. To inform clinicians about optimal rehabilitation prescription for different PwPD, future research needs to consider using broader inclusion criteria to enable adequate predictions of exercise response in this population.
The findings from this study indicate that PwPD with overall lower self-perceived health, as well as physical and cognitive ability, are likely to benefit most from highly challenging gait and balance training that is supervised. These results may suggest that PwPD with mild to moderate disease severity and overall lower levels of function should be offered to participate in challenging exercise interventions delivered at a high dosage.
The authors thank all the PwPD who participated in this study, as well as all trainers and testers. This study was funded by The Swedish Research Council, Karolinska Institutet, Neuro Sweden, the Swedish Parkinson Association, the Swedish Research Council for Health, Working Life and Welfare (FORTE), and the Vårdal foundation.
1. Hirsch MA, Iyer SS, Sanjak M. Exercise-induced neuroplasticity in human Parkinson's disease: what is the evidence telling us? Parkinsonism Relat Disord. 2016;22(suppl 1):S78–S81.
2. Abbruzzese G, Marchese R, Avanzino L, Pelosin E. Rehabilitation
for Parkinson's disease: current outlook and future challenges. Parkinsonism Relat Disord. 2016;22(suppl 1):S60–S64.
3. Tomlinson CL, Patel S, Meek C, et al Physiotherapy versus placebo or no intervention in Parkinson's disease. Cochrane Database Syst Rev. 2012(7):CD002817.
4. Shen X, Wong-Yu IS, Mak MK. Effects of exercise on falls, balance, and gait ability in Parkinson's disease: a meta-analysis. Neurorehabil Neural Repair. 2016;30(6):512–527.
5. Fox SH, Katzenschlager R, Lim SY, et al The movement disorder society evidence-based medicine review update: treatments for the motor symptoms of Parkinson's disease. Mov Disord. 2011;26(suppl 3):S2–S41.
6. Kadastik-Eerme L, Rosenthal M, Paju T, Muldmaa M, Taba P. Health-related quality of life in Parkinson's disease: a cross-sectional study focusing on non-motor symptoms. Health Qual Life Outcomes. 2015;13:83.
7. Vu TC, Nutt JG, Holford NH. Progression of motor and nonmotor features of Parkinson's disease and their response to treatment. Br J Clin Pharmacol. 2012;74(2):267–283.
8. Lewis SJ, Foltynie T, Blackwell AD, Robbins TW, Owen AM, Barker RA. Heterogeneity of Parkinson's disease in the early clinical stages using a data driven approach. J Neurol Neurosurg Psychiatry. 2005;76(3):343–348.
9. Allen NE, Sherrington C, Suriyarachchi GD, Paul SS, Song J, Canning CG. Exercise and motor training in people with Parkinson's disease: a systematic review of participant characteristics, intervention delivery, retention rates, adherence, and adverse events in clinical trials. Parkinsons Dis. 2012;2012:854328.
10. Rafferty MR, Prodoehl J, Robichaud JA, et al Effects of 2 years of exercise on gait impairment in people with Parkinson disease: the PRET-PD randomized trial. J Neurol Phys Ther. 2017;41(1):21–30.
11. Goodwin VA, Richards SH, Taylor RS, Taylor AH, Campbell JL. The effectiveness of exercise interventions for people with Parkinson's disease: a systematic review and meta-analysis. Mov Disord. 2008;23(5):631–640.
12. Strouwen C, Molenaar EALM, Münks L, et al Training dual tasks together or apart in Parkinson's disease: results from the DUALITY trial. Mov Disord. 2017;32(8):1201–1210.
13. Boyd LA, Vidoni ED, Siengsukon CF. Multidimensional motor sequence learning is impaired in older but not younger or middle-aged adults. Phys Ther. 2008;88(3):351–362.
14. Paul SS, Dibble LE, Peterson DS. Motor learning in people with Parkinson's disease: implications for fall prevention across the disease spectrum. Gait Posture. 2018;61:311–319.
15. Vaillancourt DE, Schonfeld D, Kwak Y, Bohnen NI, Seidler R. Dopamine overdose hypothesis: evidence and clinical implications. Mov Disord. 2013;28(14):1920–1929.
16. Allen NE, Sherrington C, Paul SS, Canning CG. Balance and falls in Parkinson's disease: a meta-analysis of the effect of exercise and motor training. Mov Disord. 2011;26(9):1605–1615.
17. Seidler RD, Bo J, Anguera JA. Neurocognitive contributions to motor skill learning: the role of working memory. J Mot Behav. 2012;44(6):445–453.
18. Pigott K, Rick J, Xie SX, et al Longitudinal study of normal cognition in Parkinson disease. Neurology
19. Spildooren J, Vercruysse S, Desloovere K, Vandenberghe W, Kerckhofs E, Nieuwboer A. Freezing of gait in Parkinson's disease: the impact of dual-tasking and turning. Mov Disord. 2010;25(15):2563–2570.
20. Gatchel RJ, Neblett R, Kishino N, Ray CT. Fear-avoidance beliefs and chronic pain. J Orthop Sports Phys Ther. 2016;46(2):38–43.
21. Ellis T, Boudreau JK, DeAngelis TR, et al Barriers to exercise in people with Parkinson disease. Phys Ther. 2013;93(5):628–636.
22. Afshari M, Yang A, Bega D. Motivators and barriers to exercise in Parkinson's disease. J Parkinsons Dis. 2017;7(4):703–711.
23. Ellis T, Cavanaugh JT, Earhart GM, et al Factors associated with exercise behavior in people with Parkinson disease. Phys Ther. 2011;91(12):1838–1848.
24. Conradsson D, Löfgren N, Nero H, et al The effects of highly challenging balance training in elderly with Parkinson's disease: a randomized controlled trial. Neurorehabil Neural Repair. 2015;29(9):827–836.
25. Horak FB, Mancini M, Carlson-Kuhta P, Nutt JG, Salarian A. Balance and gait represent independent domains of mobility in Parkinson disease. Phys Ther. 2016;96(9):1364–1371.
26. Conradsson D, Löfgren N, Ståhle A, Hagströmer M, Franzén E. A novel conceptual framework for balance training in Parkinson's disease-study protocol for a randomised controlled trial. BMC Neurol. 2012;12:111.
27. Hughes AJ, Daniel SE, Kilford L, Lees AJ. Accuracy of clinical diagnosis of idiopathic Parkinson's disease: a clinico-pathological study of 100 cases. J Neurol Neurosurg Psychiatry. 1992;55(3):181–184.
28. Hoehn MM, Yahr MD. Parkinsonism: onset, progression and mortality. Neurology
29. Folstein MF, Folstein SE, McHugh PR. “Mini-Mental State.” A practical method for grading the cognitive state of patients for the clinician. J Neurol Neurosurg Psychiatry. 1975;12(3):189–198.
30. Tombaugh TN, McIntyre NJ. The Mini-Mental State Examination: a comprehensive review. J Am Geriatr Soc. 1992;40(9):922–935.
31. 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.
32. Löfgren N, Lenholm E, Conradsson D, Ståhle A, Franzén E. The Mini-BESTest—a clinically reproducible tool for balance evaluations in mild to moderate Parkinson's disease? BMC Neurol. 2014;14:235.
33. Löfgren N, Benka Wallen M, Sorjonen K, Conradsson D, Franzén E. Investigating the Mini-BESTest's construct validity in elderly with Parkinson's disease. Acta Neurol Scand. 2017;135(6):614–621.
34. Lindemann U, Najafi B, Zijlstra W, et al Distance to achieve steady state walking speed in frail elderly persons. Gait Posture. 2008;27(1):91–96.
35. Tomlinson CL, Stowe R, Patel S, Rick C, Gray R, Clarke CE. Systematic review of levodopa dose equivalency reporting in Parkinson's disease. Mov Disord. 2010;25(15):2649–2653.
36. Martinez-Martin P, Gil-Nagel A, Gracia LM, Gomez JB, Martinez-Sarries J, Bermejo F. Unified Parkinson's Disease Rating Scale characteristics and structure. The cooperative multicentric group. Mov Disord. 1994;9(1):76–83.
37. Brandler TC, Oh-Park M, Wang C, Holtzer R, Verghese J. Walking while talking: investigation of alternate forms. Gait Posture. 2012;35(1):164–166.
38. Ware JE Jr, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care. 1992;30(6):473–483.
39. Yardley L, Beyer N, Hauer K, Kempen G, Piot-Ziegler C, Todd C. Development and initial validation of the Falls Efficacy Scale-International (FES-I). Age Ageing. 2005;34(6):614–619.
40. Baghi H, Noorbaloochi S, Moore JB. Statistical and nonstatistical significance: implications for health care researchers. Qual Manag Health Care. 2007;16(2):104–112.
41. Austin PC, Steyerberg EW. The number of subjects per variable required in linear regression analyses. J Clin Epidemiol. 2015;68(6):627–636.
42. Amboni M, Barone P, Iuppariello L, et al Gait patterns in Parkinsonian patients with or without mild cognitive impairment. Mov Disord. 2012;27(12):1536–1543.
43. Canning CG, Sherrington C, Lord SR, et al Exercise for falls prevention in Parkinson disease: a randomized controlled trial. Neurology
dual task; human movement system; neurology; postural control; rehabilitation
Supplemental Digital Content
© 2019 Academy of Neurologic Physical Therapy, APTA