G Model GAIPOS-3416; No. of Pages 11 Gait & Posture xxx (2011) xxx–xxx Contents lists available at SciVerse ScienceDirect Gait & Posture jo urnal homepage: www.elsevier.com/locate/gaitpost Review Interventions for coordination of walking following stroke: Systematic review a, b a c d Kristen L. Hollands *, Trudy A. Pelton , Sarah F. Tyson , Mark A. Hollands , Paulette M. van Vliet a School of Health, Sport and Rehabilitation Sciences, University of Salford, Frederick Rd. Campus, Salford M6 6PU, UK b .School of Psychology, College of Life and Environmental Sciences, University of Birmingham, UK c .School of Sport and Exercise Sciences, College of Life and Environmental Sciences, University of Birmingham, UK d .School of Health Sciences, University of Newcastle, Australia A R T I C L E I N F O A B S T R A C T Article history: Impairments in gait coordination may be a factor in falls and mobility limitations after stroke. Therefore, Received 5 May 2011 rehabilitation targeting gait coordination may be an effective way to improve walking post-stroke. This Received in revised form 20 July 2011 review sought to examine current treatments that target impairments of gait coordination, the Accepted 22 October 2011 theoretical basis on which they are derived and the effects of such interventions. Few high quality RCTs with a low risk of bias specifically targeting and measuring restoration of Keywords: coordinated gait were found. Consequently, we took a pragmatic approach to describing and quantifying Stroke the available evidence and included non-randomised study designs and limited the influence of Gait heterogeneity in experimental design and control comparators by restricting meta-analyses to pre- and Coordination Rehabilitation post-test comparisons of experimental interventions only. Results show that physiotherapy interventions significantly improved gait function and coordination. Interventions involving repetitive task-specific practice and/or auditory cueing appeared to be the most promising approaches to restore gait coordination. The fact that overall improvements in gait coordination coincided with increased walking speed lends support to the hypothesis that targeting gait coordination gait may be a way of improving overall walking ability post-stroke. However, establishing the mechanism for improved locomotor control requires a better understanding of the nature of both neuroplasticity and coordination deficits in functional tasks after stroke. Future research requires the measurement of impairment, activity and cortical activation in an effort to establish the mechanism by which functional gains are achieved. ß 2011 Elsevier B.V. All rights reserved. 1. Background forces between the paretic and non-paretic limbs [5], step lengths, widths and stance and swing phase durations [6–8]. All of these To restore efficient, independent functional walking, the impairments in gait coordination underlie overall decreased relationship of movements between body segments needs to be walking speed and endurance [9]. adaptable to allow variations to the walking pattern according to If gait coordination is impaired, then the ability to alter the basic environmental demands; such as changes in walking speed or gait pattern in response to environmental demands may also be direction and obstacle avoidance. This means that rehabilitation impaired [10]. Indeed, restoration of gait adaptability has been must restore a coordinated gait pattern which may be defined as associated with lower falls risk [11] and reduced motor coordina- ‘‘An ability to maintain a context-dependent and phase-dependent tion is correlated with poorer motor recovery [12]. Thus, poor cyclical relationship between different body segments or joints in coordination of body segments (such as the trunk, pelvis or lower both spatial and temporal domains.’’ [1] However, stroke survivors limbs) may be a mechanism underlying the frequent falls reported demonstrate many deficits of gait coordination which persist by stroke survivors [13,14] and, rehabilitation targeting impair- despite rehabilitation, including altered temporal and spatial ments in gait coordination may be a way to improve walking coordination between the head, trunk and pelvis [2,3] and ability post-stroke. impaired pelvic, knee, and ankle control during stance and Several systematic reviews [15–17] examining the effectiveness impaired excursions and timing during swing phase [4]. Poor of individual treatment approaches on gait speed have concluded interlimb coordination is reflected by asymmetries in propulsive there is insufficient evidence on which to determine the effective- ness of these approaches. This review, therefore, seeks to identify all interventions which aim to address gait coordination (as a putative * Corresponding author. Tel.: +44 0161 295 3238. mechanism of gait dysfunction) in stroke survivors and explore their E-mail address: [email protected] (K.L. Hollands). effects on gait coordination and overall gait ability. This will facilitate 0966-6362/$ – see front matter ß 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.gaitpost.2011.10.355 Please cite this article in press as: Hollands KL, et al. Interventions for coordination of walking following stroke: Systematic review. Gait Posture (2011), doi:10.1016/j.gaitpost.2011.10.355 G Model GAIPOS-3416; No. of Pages 11 2 K.L. Hollands et al. / Gait & Posture xxx (2011) xxx–xxx an examination of the hypothesis that restoration of a coordinated more complex treatment; provided the design yielded evidence for gait pattern is a mechanism to improve overall walking ability, as the potential efficacy of the intervention. For example, repeated well as to identify the gaps in knowledge and interventions which trials of walking overground were considered locomotor practice show promise for further study. and studies involving such a manipulation/intervention were included. 2. Methods of the review 2.3.3. Comparisons The aims of this review were to: Studies that compared measurement of an intervention with a control, be that of a group or condition, were included. This (1) Identify current non-surgical and non-pharmacological reha- included randomised and quasi-randomised controlled trials, bilitation interventions used to treat gait coordination impair- case–control studies, cohort studies, and before–after studies. ments after stroke. Studies which only reported data for comparisons between healthy (2) Determine the effects of identified interventions on measures control groups and stroke participants or had no control or baseline of gait coordination and overall walking ability compared with comparisons were excluded. no treatment. 2.3.4. Outcome measure In order to identify gaps in knowledge base and directions for Studies were required to include at least one measure of gait co- future research, the theoretical basis for interventions is discussed. ordination, defined as the following kinematics: spatial and Evidence for the hypothesis that improved gait coordination may temporal indices of gait symmetry, cross correlations of lower facilitate improved overall walking ability is also examined. limb or axial segment displacement, velocity or acceleration trajectories or measures of relative phase (see [1] for further 2.1. Databases searched details). Studies were excluded if outcomes were presented as means averaged across limbs and hence no indication of interlimb The following databases were searched: coordination could be extracted. Gait speed was extracted to reflect overall function, but studies which reported gait speed as Cochrane Stroke Group Trials Register. the only indicator of coordination were excluded as the informa- Cochrane Central Register of Controlled Trials (CENTRAL) (The tion gait speed provides regarding underlying impairments is Cochrane Library). limited [4,19–21]. Studies that did not explicitly state the objective MEDLINE (1950 to present). was to determine the effects on gait coordination, but used a EMBASE (1980 to present). coordination measure as an outcome measure were included. CINAHL (1982 to present). AMED (1985 to present). 2.4. Identification of relevant trials National Institutes of Health (NIH) Clinical Trials Database host: NIH (http://clinicaltrials.gov/ct). Selection of studies was performed according to Cochrane National Institute of Clinical Studies (http://www.med.mona- Review guidelines [22]. Two of the authors (KH, PvV) read each sh.edu.au/healthservices/cce/index.html). reference and, based on the inclusion criteria, independently ranked these as ‘possibly relevant’, or ‘definitely irrelevant’. If both The following rehabilitation specific databases which include identified a trial as ‘definitely irrelevant’ we excluded it. Consensus both peer reviewed and some un-peer reviewed content were was achieved through discussion, including a third author if searched: necessary (TP). Otseeker (http://www.otseeker.com/). 2.5. Data analysis OT Search (http://www.aota.org/otsearch/index.asp). Physiotherapy Evidence database (PEDro, http://www.pedro.fh- KH and PvV independently assessed the methodological quality s.usyd.edu.au/index.html). of selected studies using a modified version of the Joanna Briggs Chartered Society of Physiotherapy Research Database. Institute (JBI) critical appraisal checklist for cohort/case control REHABDATA (http://www.naric.com/research/rehab/default.cfm). [23]. For rigour and detail, additional questions from Downs and Black’s checklist [24] were added. The authors then independently