Gait impairments in Parkinson's disease (PD) are exacerbated under dual-task conditions requiring the simultaneous performance of cognitive or motor tasks. Dual-task walking deficits impact functional mobility, which often requires walking while performing concurrent tasks such as talking or carrying an object. The consequences of gait impairments in PD are significant and include increased disability, increased fall risk, and reduced quality of life. However, effective therapeutic interventions for dual-task walking deficits are limited. The goals of this narrative review are to describe dual-task walking deficits in people with PD, to discuss motor and cognitive factors that may contribute to these deficits, to review potential mechanisms underlying dual-task deficits, and to discuss the effect of therapeutic interventions on dual-task walking deficits in persons with PD.
Gait impairments and walking limitations are common among people with Parkinson’s disease (PD). While gait abnormalities are not pronounced in the early stages of PD, their prevalence and severity increase with disease progression. Within 3 years of diagnosis, over 85% of people with clinically probable PD develop gait problems [
Mobility in daily life frequently requires walking while performing simultaneous cognitive or motor tasks, such as talking with a friend or carrying a cup of coffee. Gait impairments in people with PD are exacerbated under such dual-task conditions. In recent years, dual-task walking research has expanded rapidly. The association of gait impairments with adverse consequences like increased fall risk has motivated research into clinical strategies to assess and treat dual-task walking deficits in PD. Several recent review papers have been published on dual-task posture and gait deficits among older adults and in a general neurologic population [
Single-task gait impairments in PD include reduced speed and stride length and increased double limb support time and stride-to-stride variability [
Table
Summary of studies examining dual-task walking in people with PD. Relevant individual, task, and environmental aspects of each study are included. Dual-task costs for walking and the concurrent task are included where they could be calculated.
Individual Characteristics | Task and Environmental Characteristics | Results | ||||||
Study |
Age |
Disease |
Cog. | Walking task & |
Concurrent task | Instruct. | Walking DTC | Concurrent |
| ||||||||
Bond and Morris, 2000 |
65 (10) | Webster |
STMS: |
Walk 10 m |
Motor: (1) carry tray; |
No prioritization | Speed: (1) −2%; (2) −11% |
— |
| ||||||||
Brown et al., |
67 (7) | UPDRS: |
MMSE: |
Walk 10 m (self- |
Cognitive: serial-3 |
None |
Speed: −20% |
— |
| ||||||||
Camicioli et |
67 (9) |
UPDRS: |
MMSE: |
Walk 4.6 m, turn |
Cognitive: verbal fluency (recite male names) | Not specified | Steps: −40% |
— |
| ||||||||
Campbell et |
74 (7) | H&Y: |
— | Timed Up and Go (3 m; |
Cognitive: (1) repeat |
Not specified | Steps: (1) +1%; (2) −13% |
— |
| ||||||||
Galletly and Brauer, 2005 [ |
65 (10) | UPDRS: |
MMSE: |
Walk 10 m |
Motor: (1) button press |
“Concentrate |
Speed: (1) −7%; |
Motor: |
| ||||||||
Hackney and |
65 (10) | UPDRS: |
— | Walk 5 m |
Cognitive: mental |
Not specified |
*Speed: −33% |
— |
| ||||||||
Hausdorff et al., 2003 [ |
Range: 52–82 | UPDRS: 14 H&Y: 3.1 | MMSE: 27 | Walk 20 m (normal pace; level ground) | Cognitive: serial-7 subtraction | “Walk while performing subtractions” | Stride time: −10% Stride time variability: −154% | — |
| ||||||||
LaPointe et al., 2010 [ |
67 | H&Y: 2.4 (7) | DRS-2: 136 (7) | Walk 4.3 m | Cognitive: (1) count by 1’s; (2) serial-3 subtraction; (3) recite alpha-numeric sequence | Not specified |
**Speed: (1) −2%; (2) −12%; (3) −19% |
— |
| ||||||||
Lord et al., 2010 [ |
71 (7) |
UPDRS: 39 (15) | MMSE: 27 (3) | Stand from a chair, walk 5–11 m (preferred speed; examined in home & distance varied by home) | Motor: (1) carry tray with 2 beakers of water |
“Concentrate equally on walking and task(s)” | Speed: (1) −24%; |
— |
| ||||||||
Lord et al., 2011 [ |
69 (7) |
UPDRS: |
MMSE: 28 (2) | Walk 6 m, turn 180°, walk 6 m (examined in home) | Motor: carrying a tray with 2 cups of water | “Concentrate on task as a whole” | Speed: −13% Stride time Variability: −8% |
— |
| ||||||||
O’Shea et al., 2002 [ |
68 (7) | Modified Webster Scale:12 (6) | STMS: |
Walk 10 m (preferred pace) | Motor: (1) coin transfer |
Not specified | Speed: (1) −18%; (2) −19% |
Motor: (1) −17.4% |
| ||||||||
Plotnik et al., 2009 [ |
72 (7) | UPDRS: |
MMSE: 28 (1) | Walk 2 min in a level, 25 m corridor (comfortable pace) | Cognitive: serial-7 subtraction | No prioritization | Phase coordination index: −47% | — |
| ||||||||
Plotnik et al., 2011 [ |
66 (7) | UPDRS: 35 (10) |
MMSE: 29 (1) | Walk ~80 m in a level, ~20 m corridor (comfortable pace) | Cognitive: (1) serial-3 subtraction; |
Not specified | Speed: (1) −17%; (2) −23% Stride length: (1) −11%; (2) −15% Stride time variability: (1) −39%; (2) −51% | — |
| ||||||||
Rochester et al., 2004 [ |
65 (8) | H&Y: |
MMSE: 27 (2) | Stand from a chair, walk 6.6 (1.5) m, return (preferred speed, examined in home & distance varied by home) | Motor: (1) carrying tray with 2 cups of water; Cognitive: (2) autobiographical memory task;(3) Motor + Cognitive | Not specified | Speed: (1) −9%; (2) −21%; (3) −23% |
— |
| ||||||||
Rochester et al., 2008 [ |
67 (8) | UPDRS: |
MMSE: 28 (2) | Walk 6 m, turn 180°, walk 6 m (preferred pace) | Motor: carrying a tray with 2 cups of water | “Concentrate equally on all tasks” | Speed: −13% | — |
| ||||||||
Spildooren et al., 2010 [ |
||||||||
Freezers ( |
69 (7) |
UPDRS: 38 (14) H&Y: 2.5 (0.5) | MMSE: 28 (1) | Walk 5 m: (1) straight; (2) turn 180°; (3) turn 360° | Cognitive: color identification (auditory attentional task) | No prioritization | Steps: (1) −25%; (2) −16%; (3) −13% |
— |
Non-freezers ( |
67 (7) |
UPDRS: 34 (10) H&Y: 2.4 (0.3) | MMSE: 29 (1) | Steps: (1) −7%; (2) +1%; (3) −2% |
— | |||
| ||||||||
Yogev et al., 2005 [ |
71 (8) | UPDRS: 18 (8) H&Y: 2.3 (0.4) | MMSE: 28 (2) | Walk 2 min in a level, 25 m corridor (comfortable pace) | Cognitive: (1) listen to a tape & answer questions; (2) above task + phoneme monitoring; |
No prioritization | Speed: (1) −10%; (2) −13%; (3) −19%Stride time variability: (1) −1%; (2) −6%; (3) −27% | Cognitive: (1) −42% |
| ||||||||
Yogev et al., 2007 [ |
72 (7) | UPDRS: |
MMSE: 28 (1) | Walk 2 min in a level, 25 m corridor(comfortable pace) | Cognitive: serial-7 subtraction | No prioritization | Gait asymmetry: −43% | — |
*: data collapsed across forward & backward walking; **: data collapsed across PD & control groups; Cog.: cognitive status; DLS: double limb support; DRS-2: Dementia Rating Scale-2; DTC: dual-task cost; H&Y: Hoehn & Yahr stage; Instruct.: instructions provided during dual-task conditions; MMSE: Mini-Mental State Examination; MoCA: Montreal Cognitive Assessment; STMS: Short Test of Mental Status; Steps: number of steps required to complete walking task; Time: time required to complete walking task; UPDRS: Unified Parkinson Disease Rating Scale, motor examination.
Studies of dual-task walking in PD vary substantially with respect to participant characteristics. Dual-task walking deficits increase with age among healthy adults [
Dual-task studies in PD also vary in terms of walking and concurrent task characteristics. Most examined walking on a level surface at a self-selected speed, but some included more complex walking tasks. For example, some walking tasks involved sit-to-stand transfers and/or turning [
Typically, no specific instructions are provided regarding which task to prioritize during dual-task conditions. In most cases, participants were either instructed to focus on both tasks or instructions were not specified. However, most studies quantified dual-task changes in walking only and did not measure concurrent task performance, making it difficult to determine if there were between-task trade-offs. DTCs provide a means to assess trade-offs between walking and concurrent task performance [
Studies that systematically manipulate environmental factors to determine the effects on dual-task walking deficits in PD are lacking. Most research was conducted in a clinical or laboratory environment, but some was conducted in participants’ homes [
In summary, the literature as a whole confirms the presence of significant dual-task walking deficits among persons with PD, despite methodological variations in participant characteristics, task demands, and environmental constraints. The extent of these deficits appears to vary as a function of individual, task, and environmental characteristics, but the relative contribution of each factor is not well understood. Carefully controlled studies are needed to better quantify how these factors impact dual-task walking deficits in people with PD.
It is not clear how motor and cognitive symptoms contribute to either single-task or dual-task walking deficits in PD. The motor phenotype of PD is heterogeneous, with cardinal features of rigidity, tremor, and bradykinesia [
Single-task walking deficits have been associated with a variety of motor symptoms in PD. For example, increased axial rigidity is associated with poorer performance on single-task measures of balance and functional mobility [
Several motor factors are associated with dual-task walking deficits in PD. Dual-task gait speed has been associated with disease severity, as measured by Hoehn and Yahr stage [
PD is associated with a variety of cognitive impairments, including executive function, attention, memory, language, and visuospatial impairments [
Specific cognitive functions, such as set shifting, divided or alternating attention, and response inhibition, may be particularly relevant to dual-task walking [
Cognitive impairments can contribute to dual-task walking deficits in various ways. First, they may limit the ability to compensate for gait impairments using cognitive strategies. People with PD are often taught conscious strategies to improve their gait pattern, such as focusing on walking with longer steps. The type and severity of cognitive impairments may limit the ability to use such strategies to compensate for gait abnormalities. Also, impaired executive function might result in the inappropriate or unsafe prioritization of tasks when walking under dual-task conditions. Bloem and colleagues have proposed that increased fall risk in people with PD may result in part from a “posture second” prioritization strategy, in which concurrent tasks are prioritized above walking [
The mechanisms responsible for interference between walking and concurrent cognitive or motor tasks in people with PD are not clear. Because multiple factors contribute to dual-task walking deficits, it is likely that a number of different mechanisms contribute to these deficits. In addition, characteristics of the concurrent task, such as type, domain, and difficulty, will impact the mechanisms and resources involved in dual-task performance. This section will review both nonspecific mechanisms proposed to explain dual-task interference across populations as well as specific mechanisms that may contribute to dual-task walking deficits in PD.
Two general theoretical frameworks have been proposed to explain dual-task interference. Capacity theory conceptualizes the information processing needed for dual-task performance as a flexible but limited resource [
A second general theory to explain dual-task interference is the bottleneck theory [
Several mechanisms specific to PD may also contribute to dual-task walking deficits. These mechanisms are not mutually exclusive, but might overlap with one another. Consistent with the capacity theory, a first specific mechanism in people with PD is reduced movement automaticity. Automaticity refers to the ability to perform a skilled movement without conscious or executive control or attention directed toward the movement [
A second mechanism that could contribute to dual-task walking deficits in PD is dopamine-mediated dysfunction of the basal ganglia. Multiple parallel pathways through the basal ganglia subserve different functions, including motor, cognitive, and limbic functions [
A third mechanism that could contribute to dual-task walking deficits in PD is the presence of nondopaminergic pathology, which may affect both gait and cognition. It is increasingly appreciated that the pathology of PD is not limited to dopamine but includes other neurotransmitter systems, such as serotonin, norepinephrine (noradrenaline), or acetylcholine [
In summary, research suggests a number of general and specific mechanisms that may contribute to dual-task walking deficits in PD. These mechanisms are not mutually exclusive, and the relative contribution of each may depend on factors like the symptom profile of the individual and the specific task combination performed under dual-task conditions. A better understanding of the mechanisms responsible for dual-task walking deficits in PD can inform novel therapeutic approaches and enhance our ability to identify optimal interventions.
The effects of various interventions on single-task walking in PD have been well described, but there is less research examining the efficacy of different pharmacological, surgical, or rehabilitative therapies on dual-task walking in this population. Because gait impairments in PD are exacerbated by dual-task conditions, which are common in daily life, it is important to understand how various therapeutic interventions affect dual-task walking.
The reported effects of anti-parkinson medications on walking in PD are variable, even under single-task conditions. Medications improve aspects of single-task walking, including gait speed and stride length, but may not influence others, like stride-to-stride variability [
The reported effects of surgery on single-task walking are inconsistent. For example, initial improvements in postural control and gait as a result of deep brain stimulation are not sustained beyond 2–9 years [
There is considerable research demonstrating training-related improvements in single-task walking in persons with PD [
External visual, auditory, or somatosensory cues improve both single- and dual-task walking in PD [
Cognitive or attentional strategies (e.g., focusing attention on walking with long steps) can also improve walking in people with PD [
Recent intervention studies have combined dual-task gait training with cognitive strategies to direct attentional focus and task prioritization. Even people with early PD report the need to monitor and consciously correct walking deficits [
One of the limitations in the research on dual-task walking interventions is the lack of consistent and validated measures of dual-task walking performance. Appropriate outcome measures are necessary to determine if a person with PD has dual-task walking deficits and if a given intervention effectively improves these deficits. A variety of tests, including the Stops Walking When Talking test or the Walking and Remembering Test, have been used to assess dual-task walking performance in older adults [
Research supports the efficacy of rehabilitative interventions, including external cueing, cognitive strategies, and dual-task gait training, to improve dual-task walking deficits in PD. Emerging research is examining additional treatment approaches to improve dual-task walking. For example, treadmill training with virtual reality, designed to incorporate more complex task and environmental conditions, has been shown to improve both single- and dual-task walking in people with PD [
This paper has reviewed basic and applied research related to dual-task walking deficits in people with PD. Gait impairments under both single-task and dual-task conditions are prevalent in people with PD and are associated with serious consequences. The severity of dual-task walking deficits appears to vary as a function of individual, task, and environmental characteristics, though the relative impacts of each factor are not well understood. Both motor and cognitive impairments have been associated with dual-task walking deficits in persons with PD. However, because the clinical profile of PD is heterogeneous, further research is needed to elucidate the relative contributions of each of these impairments to dual-task walking deficits. A number of general and specific mechanisms may underlie dual-task walking deficits in PD. The role of each is not clear, but might depend on the dual-task combination performed. These mechanisms inform a number of therapeutic interventions. Rehabilitation interventions, including external cues, cognitive strategies, and dual-task gait training, appear to be effective in reducing dual-task walking deficits in PD. However, a better understanding of the individual, task, and environmental factors that influence dual-task walking deficits is critical to refine existing interventions and identify novel therapeutic approaches.
This work was supported by the National Institutes of Health, National Institute of Child Health and Human Development (K01HD052018).