Secondary Logo

Journal Logo


Training with Verbal Instructional Cues Results in Near-term Improvement of Gait in People with Parkinson Disease

Lehman, David A. , PhD, PT1; Toole, Tonya , PhD2; Lofald, Dan , PhD3; Hirsch, Mark A. , PhD4

Author Information
Journal of Neurologic Physical Therapy: March 2005 - Volume 29 - Issue 1 - p 2-8
doi: 10.1097/
  • Free



Parkinson disease (PD) is caused by the degeneration of the dopaminergic nigrostriatal neurons of the basal gan-glia.1 Gait disturbances are among the primary symptoms of PD and can lead to falls.2,3 The resulting fear of falls can lead to decreased activity and further loss of independence.4

The basal ganglia are thought to be involved in the execution of repetitive and automatic movements, such as walking, and are believed to be one of the main sources of dysfunctional movement execution in PD.5–8 Morris and colleagues hypothesized that the basal ganglia are involved in providing internal phasic cues to the supplemental motor area (SMA) and assist in the transmission of motor-set information.9 According to this model, the basal ganglia activate and deactivate submovements of a movement sequence and are responsible for the accurate execution of each sub-movement element and the suppression of undesired movement programs.10–14

Damage to the basal ganglia may result in ‘choppy’ move-ment execution due to improper supply of internal rhythmic cues and abnormalities in the actual movement ele-ment.5,9,15,16 According to Morris and colleagues, 2 of the features of gait in PD—freezing and festination—may be due to dysfunctional internal rhythmic cues.9 Deficits in motor set could produce abnormally short, shuffling steps, and reduced or absent arm swing.

Damage to the basal ganglia could interfere with motor learning, defined as a relatively long-term change in the ability to perform an action. Therapeutic interventions that focus on teaching gait strategies to facilitate motor learning through external auditory, visual, or attentional cueing among persons with PD could potentially improve gait and thereby contribute to a more active and independent lifestyle in these persons.

Motor learning studies show that persons with PD have difficulty in initiating and maintaining movement without external cues.17 Cueing strategies are thought to reroute the movement through a nonautomatic pathway, removing it from the automatic basal ganglia pathway. For example, studies using external rhythmic auditory cues to supplement an absent or deficient internal rhythm in PD have shown changes in gait velocity, cadence, and stride-length.18–21 Studies using visual cues such as floor markers to augment the defective motor set, or a combination of visual and auditory cues have generally shown similar effects on gait.9,22–26 In contrast, studies have rarely used verbal instructional cues to focus attention on walking and studies have not examined the long-term effects of instructional cues on gait in PD.27

The aim of this study was to assess the immediate and near-term effects of a verbal instructional set on select gait patterns in persons with PD. Persistent effects from such a training program would support the notion that motor learning in individuals with early stage PD can occur with practice. The study was executed in 2 parts. Part one was a within subject single group design and part 2 was a between group (treatment and control design).

We hypothesized, for part one, that after completion of a 10-day training program using instructional cues, there would be improvement of gait parameters immediately after the training program, one week after the cessation of the training program, and one month after the cessation of the training program (as compared to pretraining values). We hypothesized for part 2, that there would be measurable differences of the post-training gait values between the control (nontraining) group and the treatment group.


Part One


A sample of convenience of 5 individuals (3 male and 2 female) with gait impairments due to Parkinson Disease were recruited from support groups, physician offices, and newspaper announcements (Table 1). Persons with other neurological and/or orthopaedic impairments that could not walk the distances required of the training program were excluded. The human subjects review board at the Florida State University and the University of St. Augustine for Health Sciences approved the study and participants gave their informed consent. Impairment of gait was rated according to the Unified Parkinson Disease Rating Scale (UPDRS). The UPDRS was conducted at the time of day the participant reported her/his medication to be most effective (ie, 1 hour after administration of L-Dopa). All participants were similar in their gait deficit measures according to the Motor Examination sections of the UPDRS (Table 2). All participants were classified as Hoehn & Yahr Stage 2–2.5 and fell within their normative range on mentation, behavior, and mood as determined by the Mini Mental State Exam (MMSE).28

Table 1:
Characteristics of Participants
Table 2:
UPDRS Scores for Each Section


The GAITRite® (CRI Systems, Clifton, NJ) was used to assess gait parameters of step length, velocity, and cadence. The GAITRite is an electronic walkway with encapsulated sensor pads. Values from the GAITRite have been compared with other physical and recording measures (concurrent validity) and have returned interclass coefficients greater that 0.95.29 The GAITRite was further compared against a robotic stride simulator that returned exact stride lengths. The step length of the simulator was set at 64.8 ± 0.0 cm and 49.4 ± 0.00 cm. The measurements taken with the GAITRite from the center of the walkway were 64.9 ± 0.6 cm and 49.3 ± 0.3 cm. It was concluded that the GAITRite returns both valid and reliable measures of stride length.

Experimental controls

For this study, the GAITRite walkway was centered in the middle of a 30-foot pathway, with ample room around all borders of the pathway. The area before and after the GAITRite walkway was of the same material as the GAITRite walkway. This allowed the participant to have a smooth transition to a steady state gait before data acquisition.

Participants were scheduled for testing and treatment at the same time each day. Any participant taking medication for PD participated in the sessions 1 to 2 hours after taking the medication. The participants wore loose fitting clothes that would not impair their ability to walk. The participants all wore flat bottom shoes.


Training phase

The participant's training session was at the same time each day (1–2 hours after taking Parkinson disease medication). One trip down the 30-foot pathway is a length. Each training set consisted of 20 lengths. Participants completed 3 training sets each day. In total, participants walked 1800 feet per day. The training phase lasted for 10 days (Monday through Friday for 2 consecutive weeks).

The instructional set, “take long steps,” was stated at the beginning of the first set and every 2 lengths thereafter until the 20-length set was completed. The instructional set was reinforced 10 times per training session, totaling 300 repetitions for the intervention phase of the experiment. There was a 5-minute rest period between sets. There was no feedback provided to the participant concerning his/her performance.


Baseline, fifth-day, tenth-day, and post-test measurements were gathered under 2 conditions: (1) preferred walking and (2) walking with the use of the instructional set, “take long steps.” Five trials of the preferred condition preceded the 5 trials of the cued condition. This order was to prevent the instructional set from influencing the preferred walking condition.

Instructions for the preferred walking condition were, Mr/Mrs—, I want you to walk the length of this walkway when I say go. I want you to walk at your normal, most comfortable pace, just like you walk everyday. Instructions for walking with the use of the instructional set were, “Mr/Mrs—, I want you to now walk across the walkway and take long steps.” The instructions were delivered by the same person and in the same tone of voice. The instructions were repeated for each trial.

Data Analysis

A 2 (cue/no cue) × 5 (baseline, day 5 of treatment, day 10 of treatment, 1 week after treatment, and 1 month after treatment) repeated measures ANOVA was used to test for main and interaction effects. Tukey's HSD analysis was used to maintain the alpha level at the 0.05 level for all pairwise comparisons.

Part Two


In order to limit the threat to internal validity, a second part to this study required a control group to isolate the effects due to the training component of the intervention. Eleven new participants, qualifying under the same inclusion criteria as in study one, were randomly assigned to either an intervention (training) group (n = 6) or a control (nontreatment) group (n = 5). Subject characteristics for part 2 are presented in Table 3. The participants' scores on the sections of the UPDRS are presented in Table 4. As in part one, scores on the UPDRS for participants in part 2 are also indicative of early stage Parkinson disease.

Table 3:
Descriptive Characteristics of Participants
Table 4:
UPDRS Scores for Each Section


Equipment, controls (eg, not changing lifestyle or medicine), and treatment intervention were the same in the second study. The control group did not receive any form of intervention (ie, the injunction to take long steps nor the ability to practice long steps while being cued).

As a measure of the training effects, all participants were tested for preferred walking as described in part one. Measurements of step length were taken for baseline, after 10 days of training, 1 week after training, and 1 month after training. Testing instructions and conditions were the same as described in part one.


Part One

There was a significant main effect of cue on step length of the left (F (3, 29) = 2. 24, P < 0.05) and the right (F (3,29) = 2. 24, P < 0.05) lower extremity. Mean step length of the left lower extremity in the cued condition (x = 70.84 cm) was statistically larger than mean step length in the non-cued condition (x = 56. 84 cm). Mean step length of the right lower extremity in the cued condition (x = 70.38 cm) was statistically larger than mean step length in the non-cued condition (x = 57. 98 cm).

There was a significant main effect of cue on velocity (F (3,29) = 2. 24, P < 0.05). The mean velocity for the cued condition (x = 114. 506 cm/sec) was statistically larger than the mean velocity for the noncued condition (x = 99. 43 cm/sec). There was a significant main effect of cue on cadence. The mean cadence for the cued condition (x = 98. 19 steps/min) was statistically smaller than the mean cadence for the noncued condition (x = 104. 66 steps/min). There was no interaction of cue and time for any of the dependent variables.

There was a significant main effect of time (see Figure 1) for step length of the left leg (F(3,29) = 2. 24, P <. 05). The mean step length of the left lower extremity at baseline was x = 45. 96 cm. This value is statistically shorter than mean step length after 5 (x = 61. 62 cm) and 10 (x = 59. 84 cm) days of training. There were no statistical differences between day 5 and day 10. After the end of the treatment phase, step length of the left lower extremity remained significantly larger at 1 week after treatment (x = 57. 63 cm) and 1 month after treatment (x = 59. 34 cm). The same patterns of effects were found for the right lower extremity as are reported here for the left extremity.

Figure 1:
Main effect of time on step length.

There was no interaction of cue and time for any of the dependent variables. Summary tables of means and standard deviations of all significant main effects for cue and time are provided in Tables 5 and 6.

Table 5:
Summary of Means and Standard Deviations for Significant Main Effects for Time and Cue on Step Length (cm)
Table 6:
Summary of Means and Standard Deviations for Significant Main Effects for Time and Cue on Velocity (cm/sec) and Cadence (steps/min)

Part Two

Step length of the right extremity, velocity, and cadence were removed from our analysis after confirming that the relationship was isotropic in their relationship with step length of the left lower extremity. There was a significant 2-way interaction for Group by Time for Step Length, (F (3,27) = 3. 924, P = .019) (see Figure 2). Tukey post-hoc analysis revealed significant differences of Step Length for the treatment group between baseline (x = 55. 84 cm) and immediately after the 10-day training session (x = 63. 95 cm), 1 week after training (x = 63. 25 cm), and 1 month after training (x = 60.14 cm). Post-hoc analysis also revealed significant differences between the treatment group and control group at each post-test: immediately after the 10-day training session (control: x = 51. 70 cm, treatment: x = 63. 95), 1 week after training (control: x = 52. 50 cm, treatment: x = 63. 25 cm), and 1 month after training (control: x = 51. 98 cm, treatment: x = 60.14 cm).

Figure 2:
Two-way interaction of group by time.


We examined the immediate and near-term effect of a verbal instructional set (cueing) on select gait patterns in persons with PD. There were 4 main findings: (1) gait velocity and step length increased in response to cueing, (2) cadence was reduced in response to cueing, (3) these effects were observed in both the experimental and control group but the effects were more pronounced in the experimental group, (4) the effect persisted for at least 4 weeks in both groups.

The effect of training on step length, step velocity, and cadence was readily apparent immediately following train-ing. This result extends previous observations showing the immediate effects of using external cueing to improve motor function in persons with PD.9,17–21,25,27,30

In the present study, the improvements in gait were sustained throughout the follow-up period and for up to 1 month after cessation of training. This is the first study to show near-term effects of instructional cueing on gait in persons with PD. Our findings extend previous observations showing near-term effects lasting up to 4 weeks in studies using auditory cues to improve gait in persons with PD.18–21 These studies did not use a control group, making it difficult to compare results to our study. In the present study, participation in the control group (that did not receive training) resulted in a small but statistically significant improvement in gait from base line to postmeasure-ments. This suggests that practice and instruction given to participants during the base-line segment of part two may have been a contributing factor to the changes in gait observed in the control group during post-testing, and demonstrates a motor learning capacity for individuals in this group. Participation in the instructional cueing group resulted in significantly higher mean step-length and gait velocity scores and lower mean cadence scores than participation in the control group, showing the strength of the instructional cueing treatment effect on motor learning.

Although we did not assess the mechanisms for motor learning in this study, there are several possible explanations. Learning of new movement sequences has been defined to occur either implicitly or explicitly.31 Implicit learning is thought to occur through nonattentional unconscious processes, leading to procedural knowledge. Explicit learning is thought to use a conscious process requiring attention, leading to declarative knowledge.31 The evidence regarding dysfunction among either of these systems among persons with PD is controversial. Several studies have demonstrated impairment of explicit and implicit learning in persons with PD, while other studies have shown these systems to be intact in PD.31–34 Impairment in these attentional systems may be due to accelerated cognitive decline or frontal lobe pathology, which may occur in some persons with PD.35

In the present study, participants had normal cognitive test scores, indicative of no age-associated cognitive pathology. Since motor performance remained unchanged relative to the post-test and throughout the follow-up period, participants did not receive instructional cue training during the postassessment phase; and there were no instructional cues delivered to participants during all postgait measurements. The instructional cue training may have tapped into the participants', presumably intact, implicit motor learning system.

Studies also show that persons with PD require a large number of repetitions of the task to translate declarative knowledge into procedural knowledge.34 This may be relevant to our study results. Participants in this study received instructional cues for 10 days in succession and practiced their walking for 1800 feet per day. The experimental group clearly outperformed the control group on the gait measures. The high instructional cue intensity may have translated declarative knowledge during the training phase into procedural knowledge during the postmeasurements of gait.

Further evidence shows that a high number of practice sessions are a potentially crucial component of rehabilitation for PD. Since PD is a progressive neurodegenerative condition, signs and symptoms are expected to become more severe over time, even with optimal pharmacological therapy. Recent work with animal models of PD indicate that rehabilitative training can stimulate a number of plasticity-related events in the brain, including neurotrophic factor expression and synaptogenesis.36–40 Training appears to promote behavioral recovery by modulating genes and proteins important to basal ganglia function.36,40 Moreover, during slow degeneration of nigrostriatal dopaminergic neurons, coapplication of intense sensorimotor training appears to be neuroprotective.36–40

Studies examined the use of forced-use of the impaired forelimb soon after unilateral exposure to a potent dopaminergic neurotoxin (6-hydroxydopamine – 6-OHDA), which produces signs and symptoms comparable to those seen in humans with idiopathic PD.38 Forcing an animal to use its impaired forelimb, while restricting its less involved forelimb in a cast, is a relatively high intensity activity. The animal must perform a high number of repetitions of a movement with its more impaired forelimb over the course of therapy. Forced-use studies in animals produced predictable neurochemical and structural effects, such as sparing of striatal dopamine, and slowing of dopaminergic neuron degeneration.38 Interestingly, those animals forced to use their impaired forelimb early after the lesion (3 days) outperformed those who were treated later and the behavioral and neurochemical effects of the early treatment group lasted for 60 days (after which the animals were sacrificed).

The results from the animal studies may be applicable to the present study. The participants in the present study were a relatively homogenous group, presumably unaffected by overt cognitive pathology, and with Hoehn and Yahr and UPDRS scores suggestive of early disease stage. The participants practiced walking for 10 days totaling 18,000 yards, under conditions of instructional cueing. Additionally, the results from the animal studies suggest that the early higher intensity treatment of basal ganglia dysfunction may be a key to the success of interventions for persons with PD, when the nervous system is perhaps more plastic.

The findings of this study are relevant to physical therapy practice. Practitioners already incorporate verbal instructions in their interventions for individuals with PD. The results of this study may serve as guide for the frequency, intensity, and duration of verbal cues needed for the retention of gait improvements.


The results from our study provide evidence that damage to the basal ganglia need not have an impact on the capacity of the brain to learn new movement sequences, such as those involved in gait. New movement sequences may be retained for quite some time, in the absence of instructional cueing. The effect of instructional cueing on gait has rarely been studied in persons with PD and this study is the first to show near-term effects of instructional cueing (Take Long Steps) on gait in persons with PD. Although further study is necessary to establish the relationship between cues and motor learning, we hypothesize that a gait-training program involving instructional cues is an effective way to modify dysfunctional gait parameters among people with PD.


1 Marsden CD, Obeso JA. The function of the basal ganglia and the paradox of stereotaxic surgery in Parkinson's disease. Brain. 1994;117:877–897.
2 Olanow CW, Koller WC. An algorithm (decision tree) for the management of Parkinson's disease. Neurology. 1998;50(3):S157.
3 Schaafsma JD, Giladi N, Balash Y, Bartels AL, Gurevich T, Hausdorff JM. Gait dynamics in Parkinson's disease: relationship to Parkinsonian features, falls and response to levodopa. J Neurol Sci. 2003;212(1–2):47–53.
4 Friedman SM, Munoz B, West SK, Rubin GS, Fried LP. Falls and fear of falling: which comes first? A longitudinal prediction model suggests strategies for primary and secondary prevention. J Am Geriatr Soc. 2002;50(8):1329–1335.
5 Cunnington R, Iansek R, Bradshaw J, Phillips J. Movement-related potentials in Parkinson's disease: presence and predictability of temporal and spatial cues. Brain. 1995;118:935–950.
6 Georgiou N, Iansek R, Bradshaw J, Phillips J, Mattingley J, Bradshaw J. An evaluation of the role of internal cues in the pathogenesis of parkinsonian hyperkinsias. Brain. 1993;116:1575–1587.
7 Marsden CD. The mysterious motor function of the basal ganglia: the Robert Wartenberg Lecture. Neurology. 1982;32(5):514–39.
8 Morris ME, Matyas TA, Iansek R, Summers JJ. Temporal stability of gait in Parkinson's disease. Phys Ther. 1996;76:763–777.
9 Morris ME, Iansek R, Matyas TA, Summers JJ. Stride length regulation in Parkinson's disease. Normalization strategies and underlying mechanisms. Brain. 1996;119:551–568.
10 Jueptner M, Weiller C. A review of differences between basal ganglia and cerebellar control of movements as revealed by functional imaging studies. Brain. 1998;121:1437–1449.
11 Schmidt WJ. Dopamine-glutamate interactions in the basal ganglia. Amino Acids. 1998;14:5–10.
12 Kropotov JD, Etlinger SC. Selection of actions in the basal ganglia-thalamocortical circuits: review and model. Int J Psychophysiol. 1999;31:197–217.
13 Redgrave P, Prescott TJ, Gurney K. The basal ganglia: a vertebrate solution to the selection problem? Neuroscience. 1999;89:1009–1023.
14 Brooks DJ. Imaging basal ganglia function. J Anat. 2000;196:543–554.
15 Hausdorff J, Cudkowicz M, Firtion R, Wei J, Goldberger A. Gait variability and basal ganglia disorders: stride-to-stride variations of gait cycle timing in Parkinson's disease and Huntington's disease. Movement Disorders. 1997;13:428–437.
16 Giladi N, McDermont M, Fahn S, Przedborski S, Jankovic J, Stern M, Tanner C; The Parkinson Study Group. Freezing of gait in Parkinson's disease. Neurology. 2001;56:1712–1721.
17 Rogers MA, Phillips JG, Bradshaw JL, Iansek R, Jones D. Provision of external cues and movement sequencing in Parkinson's disease. Motor Control. 1998;2(2):125–132.
18 McIntosh GC, Brown SH, Rice RR, Thaut MH. Rhythmic auditory-motor facilitation of gait patterns in patients with Parkinson's disease. J Neurol Neurosurg Psychiatry. 1997;62:22–26.
19 McIntosh G, Thaut M, Rice R, Miller R. Stride frequency modulation in parkinsonian gait using rhythmic auditory stimulation. Ann Neurol. 1994;36:316.
20 Ito N, Hayshi A, Lin W, Ohkoshi N, Watanabe M, Shoji S. Music therapy in Parkinson's disease: improvement of parkinsonian gait and depression with rhythmic auditory stimulation. In: Nakada T, ed. Integrated Human Brain Science: Theory, Method Application (music). New York, NY: Elsevier Science; 2000:435–443.
21 Thaut MH, McIntosh KW, McIntosh GC, Hoemberg V. Auditory rhythmicity enhances movement and speech motor control in patients with Parkinson's disease. Funct Neurol. 2001;16(2):163–172.
22 Azulay J, Mesure S, Amblanrd B, Blin O, Sangla I, Pouget J. Visual control of locomotion in Parkinson's disease. Brain. 1999;122:111–120.
23 Morris ME, Iansek R, Matyas TA, Summers JJ. Ability to modulate walking cadence remains intact in Parkinson's disease. J Neurol Neurosurg Psychiatry. 1994;57:1532–1534.
24 Bagley S, Kelly B, Tunnicliffe N, Turnbull G, Walker J. The effect of visual cues on the gait of independently mobile Parkinson's disease patients. Physiotherapy. 1991;77:415–420.
25 Morris ME, Iansek R, Matyas TA, Summers JJ. The patho-genesis of gait hypokinesia in Parkinson's disease. Brain. 1994;117:1169–1181.
26 Morris M, McGinley J, Huxham F, Collier J, Iansek R. Constraints on the kinetic, kinematic and spatiotempo-ral parameters of gait in Parkinson's disease. Hum Mov Sci. 1999;18:461–483.
27 Behrman L, Teitelbaum P, Cauraugh JH. Verbal Instructional sets normalise the temporal and spatial gait variables in Parkinson's disease. J Neurol Neuro-surg Psychiatry. 1998;65:580–582.
28 Crum RM, Anthony JC, Bassett SS, Folstein M F. Population based normals for the mimi-mental state by age and educational level. JAMA. 1993;18:2386–2391.
29 McDonough AL, Batavia M, Chen FC, Kwon S, Ziai J. The validity and reliability of the GAITRite system's measurements: A preliminary evaluation. Ach Phys Med Rehabil. 2001;82:419–425.
30 Morris ME, Iansek R, Matayas T, Summers J. Abnormalities in the stride length-cadence relation in parkinsonian gait. Movement Disorders. 1998;13:61–69.
31 Kelly SW, Jahanshahi M, Dirnberger G. Learning of ambiguous versus hybrid sequences by patients with Parkinson's disease. Neuropsychologia. 2004;42(10): 1350–1357.
32 Jackson GM, Jackson SR, Harrison J, Henderson L, Kennard C. Serial reaction time learning and Parkinson's disease: evidence for a procedural learning deficit. Neuropsychologia. 1995;33(5):577–593.
33 Ferraro FR, Balota DA, Connor LT. Implicit memory and the formation of new associations in nondemented Parkinson's disease individuals and individuals with senile dementia of the Alzheimer type: a serial reaction time (SRT) investigation. Brain Cogn. 1993;21(2):163–180.
34 Pascual-Leone A, Grafman J, Clark K, et al. Procedural learning in Parkinson's disease and cerebellar degeneration. Ann Neurol. 1993;34(4):594–602.
35 Locascio JJ, Corkin S, Growdon JH. Relation between clinical characteristics of Parkinson's disease and cognitive decline. J Clin Exp Neuropsychol. 2003;25(1):94–109.
36 Cohen AD, Tillerson JL, Smith AD, Schallert T, Zigmond MJ. Neuroprotective effects of prior limb use in 6-hydroxydopamine-treated rats: possible role of GDNF. J Neurochem. 2003;85(2):299–305.
37 Tillerson JL, Cohen AD, Caudle WM, Zigmond MJ, Schallert T, Miller GW. Forced nonuse in unilateral parkinsonian rats exacerbates injury. J Neurosci. 2002;22(15):6790–6799.
38 Tillerson JL, Cohen AD, Philhower J, Miller GW, Zigmond MJ, Schallert T. Forced limb-use effects on the behavioral and neurochemical effects of 6-hydroxydopamine. J Neurosci. 2001;21(12):4427–4435.
39 Tillerson JL, Caudle WM, Reveron ME, Miller G W. Exercise induces behavioral recovery and attenuates neurochemical deficits in rodent models of Parkinson's disease. Neuroscience. 2003;119(3):899–911.
40 Fisher BE, Petzinger GM, Nixon K, et al. Exercise-induced behavioral recovery and neuroplasticity in the 1-methyl-4-phenyl 1,2,3,6-tetrahydropyridine-lesioned mouse basal ganglia. J Neurosci Res. 2004;77(3):378–390.

gait; step length; velocity; cadence; basal ganglia; motor learning

© 2005 Neurology Section, APTA