Effect of Rhythmic Auditory Cueing on Aging Gait: A Systematic Review and Meta-Analysis

Rhythmic auditory cueing has been widely used in gait rehabilitation over the past decade. The entrainment effect has been suggested to introduce neurophysiological changes, alleviate auditory-motor coupling and reduce cognitive-motor interferences. However, a consensus as to its influence over aging gait is still warranted. A systematic review and meta-analysis was carried out to analyze the effects of rhythmic auditory cueing on spatiotemporal gait parameters among healthy young and elderly participants. This systematic identification of published literature was performed according to PRISMA guidelines, from inception until May 2017, on online databases: Web of science, PEDro, EBSCO, MEDLINE, Cochrane, EMBASE, and PROQUEST. Studies were critically appraised using PEDro scale. Of 2789 records, 34 studies, involving 854 (499 young/355 elderly) participants met our inclusion criteria. The meta-analysis revealed enhancements in spatiotemporal parameters of gait i.e. gait velocity (Hedge’s g: 0.85), stride length (0.61), and cadence (1.1), amongst both age groups. This review, for the first time, evaluates the effects of auditory entrainment on aging gait and discusses its implications under higher and lower information processing constraints. Clinical implications are discussed with respect to applications of auditory entrainment in rehabilitation settings.

, and cognitive processing [27,28]. Reelick, et al. [23], for instance, suggested a reduction in self-confidence with aging, and history of falls often leading to a peculiar "fear of falling" [29,30]. Furthermore, this "fear" has been reported to additionally modify the stability during static, and dynamic postures [9,31,32]. Giladi, et al. [32], referred such modified gait pattern as a "cautious or fearful gait" [23]. Although these modifications are aimed to enhance stability during locomotion, they, in turn, develop a stiff, slow and unsteady gait pattern [33]. Moreover, this "fear of falling" or "cautious gait" might promote "internal" attentional focus [34], explicit motor control [25], and can eventually alleviate cognitive-motor interferences [35] (see also Young and Mark Williams [33]). Masters and Maxwell [27] suggested that such an attempt to consciously monitor or control an autonomic movement, such as posture, or gait might adversely affect its performance. Also, such higher information processing constraints have demonstrated detrimental effects on proprioceptive perceptions [36][37][38], which are integral for autonomic stability [36]. In addition, literature suggests that younger population groups, on the contrary, have a more resilient and stable psycho-physiological stature [35,39]. However, falls are not uncommon [10]. Possibly, environmental [10], and lifestyle factors might play a considerable role [40]. Schabrun, van den Hoorn, Moorcroft, Greenland and Hodges [41] reported texting and reading while walking (common among youngsters) to adversely impact gait stability [42], by increasing cognitive-motor interferences [43]. Consequently, such higher attentional constraints predisposing to falls might possess serious life-threatening consequences under "high-stress" environments [8,44], for both younger and elderly age groups.
Additionally, rhythmic auditory entrainment is cheap [77], viable [78], easy to follow and has shown enhancements even during unsupervised home-based training programs [79,80]. This intervention can be a useful rehabilitation tool in middle and lower income countries, where poor healthcare services [81], might precipitate to majority of the fall related deaths [4]. Thereby, strongly warranting the need for such economical, and efficient rehabilitation techniques.
High-quality systematic reviews and meta-analyses have been carried out to evaluate the beneficial effects of rhythmic auditory cueing on gait in patients affected from neurological conditions, such as stroke, and parkinsonism [57,58,66]. However, to the best of our knowledge, no review to date has analyzed the effects of rhythmic auditory cueing on aging gait. Therefore, we attempted to develop a state of the art knowledge for the use of rhythmic auditory cueing in gait rehabilitation across healthy population groups. The main aim of this review is to understand the effects of auditory entrainment on spatiotemporal, variability parameters for gait among young, and elderly age groups. The review also discusses possible applications of auditory entrainment in rehabilitation and activities for daily living.
measures, results, and conclusions. The data were then summarized and tabulated ( Table 1).
The inclusion criteria for the studies was (i) Performed studies were either randomized controlled trials, cluster randomized controlled trials or controlled clinical trials; (ii) Studies reporting reliable and valid spatiotemporal gait parameters (iii) Studies reporting dynamic aspects of gait stability (iv) Studies qualified PEDro methodological quality scale (≥4 score); (v) Experiments conducted on human participants; (vi) Published in a peer-reviewed academic journal; (vii) Articles published in English and German languages.

Quality & risk of bias assessment
The quality of the studies was assessed using the PEDro methodological quality scale [83]. The scale consists of 11 items addressing external validity, internal validity, and interpretability and can detect potential bias with fair to good reliability [84], and validity [83]. A blinded rating of the methodological quality of the studies was carried out by the primary reviewer. Ambiguous issues were discussed with second (IG), third (AOE) reviewer and consensus was reached. Included studies were rated, and interpreted according to scoring of 9-10, 6-8 and 4-5 considered of "excellent", "good" and "fair" quality [85], respectively. Inadequate randomization, non-blinding of assessors, no intention to treat analysis and no measurement of compliance were considered as major threats to biasing [86].

Data Analysis
This systematic review included a meta-analysis approach [87]. The presence and lack of heterogeneity asserted the use of either random or fixed effect meta-analysis [88], respectively. A narrative synthesis of the findings structured around the intervention, population characteristics; methodological quality (Table 1) and the type of outcome are provided. Likewise, summaries of intervention effects for each study are also provided in a tabular form (Table 1). A meta-analysis was conducted between pooled studies using CMA (Comprehensive meta-analysis V 2.0, USA). Heterogeneity between the studies was assessed using I 2 statistics. The data in this review was systematically distributed and for each available variable pooled, dichotomous data was analyzed and forest plots with 95% confidence intervals are plotted. The weighted effect sizes are reported as Hedge's g [89]. Thresholds for interpretation of effect sizes were as follows; a standard mean effect size of 0 means no change, negative effect size means a negative change, mean effect size of 0.2 considered a small effect, 0.5 a medium effect and 0.8 a large effect [90]. Interpretation of heterogeneity via I 2 statistics was as; 0-0%, 25%, 75% as negligible, moderate and substantial heterogeneity, respectively. Meta-analysis reports including heterogeneity among studies were evaluated to determine the reason of heterogeneity, and the included studies were then pooled separately and analyzed again. The alpha level was set at 95%.

Characteristics of included studies
Our initial search yielded a total of 2789 studies, which on implementing our inclusion/exclusion criteria, were reduced to thirty-four ( Fig. 1). Data from the included studies have been summarized in (Table 1). Of the thirtyfour included studies, one was randomized controlled trial, and thirty-three were controlled clinical trials.

Participants
A total of 854 participants were analyzed in the incorporated studies. Studies were then categorized into sub-groups for evaluating young and elderly participants. Three studies compared the effects of rhythmic auditory cueing amongst young and elderly participants. Eighteen studies evaluated elderly participants (68±5.6 years). A total of 355 participants were evaluated (235 females/ 100 males). Two studies did not specify the gender of the participants. All the studies evaluated a mixed gender sample size. Nineteen studies evaluated young participants (26.8±6 years). A total of 499 participants were evaluated (215 females/ 248 males). Two studies did not specify the gender of the included participants [91,92]. Only one study evaluated a non-mixed gender sample i.e. only females [93]. Descriptive statistics relating to the age (mean ± standard deviation) of the participants were tabulated across the studies ( Table 1).

Risk of bias
The review included studies scoring ≥4 on PEDro to reduce the incidence of biasing. Moreover, the limitation of research protocols to be included in the review was limited to gold standard randomized controlled trials, cluster randomized controlled trials and controlled clinical trials. The individual scores attained by the studies using the PEDro scale have been reported (Table 1,  Supplementary table 2). The average PEDro score for the fifty included studies was computed to be 4.7 out of 10, indicating fair-quality of the overall studies. One study scored 8, four scored 6, fourteen studies scored 5, and sixteen studies scored 4. Publication bias was analyzed by plotting a Hedge's g against standard error (Fig. 2). Asymmetries concerning mean in the funnel plot might suggest bias (either positive or negative), in which case results are published. Risk of bias across the studies has been demonstrated in Fig. 3.

Outcomes
The results suggest evidence for a positive impact of rhythmic auditory cueing on spatiotemporal gait parameters amongst both young and elderly participants.
In the included thirty-four studies, thirty studies reported significant enhancements, two studies reported enhancements (p>0.05) [94,95], and two studies reported significant reduction in gait parameters with rhythmic auditory cueing [96,97].

Meta-analysis report
The evaluation of research studies via meta-analysis requires strict inclusion criteria to efficiently limit the heterogeneity [98]. However, among the pooled group of studies post strict inclusion criteria, some amount of unexplained heterogeneity was still observed. Sub-group analysis was then performed for identical studies to evaluate the cause of heterogeneity. The evaluated parameters were the spatio-temporal gait parameters such as, cadence, stride length, gait velocity, coefficient of variability for stride time and stride length. The effects of fast/slow tempo on gait parameters in the included studies was determined by keeping the patient's preferred cadence as reference. Analyses were also conducted to evaluate the effects of dual-task conditions, presence of instructions, and different tempo at which rhythmic auditory cueing was provided on gait parameters. We included a generalized group analysis first combined for all the pooled studies. A separate analysis in addition to clinical controlled trials was performed for high quality randomized controlled trails, for allowing a better interpretation of the direction and magnitude of effects. The main reason for not including the statistical approach within the studies was due to major differences in between assessment methods and lack of descriptive statistics within the manuscript. However, data was not received even after contacting the respective corresponding authors.

Gait velocity
The meta-analysis on healthy patients revealed (Fig. 4)  The analysis for young participants performing gait with rhythmic auditory cueing revealed ( Supplementary Fig. 1) beneficial effects with large effect and substantial heterogeneity (g: 0.92, 95% C.I: 0.42 to 1.41, I 2 : 93.2%, p<0.01). Further, sub-group analysis with non-modulated rhythmic auditory cueing ( Supplementary  Fig. 2), under a single task condition, revealed a large effect size with substantial heterogeneity (Hedge's g: 1.24, 95% CI: 0.4 to 2, I 2 : 90.5%, p<0.01). The heterogeneity here could be attributed to different interventions utilized by studies. Wellner, et al. [91] for instance, utilized robot assisted gait, and Almeida, et al. [92] analyzed treadmill gait. Moreover, different measures of rhythmic auditory cueing were utilized by [99], as the study reported generation of rhythmic patterns by converting the foot strike patterns to rhythmic pattern in real-time.
Further, analysis with fast paced stimuli revealed ( Supplementary Fig. 3) large effect size with substantial heterogeneity (g: 1.17, 95% C.I: 0.38 to 1.96, I 2 : 91.4%, p<0.01). Likewise, slow paced stimuli revealed ( Supplementary Fig. 4) reduction in gait velocity parameters with medium effect size and substantial heterogeneity (g: -0.3, 95% C.I: 90.4%, I 2 : 90.4%, p<0.01). Here as well, the heterogeneity could be attributed to the type of entrainment used, for instance, low groove, non-motivating cueing and slow cueing were paired together and vice versa for the fast-paced stimuli. These stimuli differ in terms of emotional and expressiveness components, which might be considerably different from each other [68].

Stride length
The meta-analysis on healthy patients revealed (Fig. 5) a medium effect size in positive domain with substantial heterogeneity (Hedge's g: 0.61, 95% CI: 0.23 to 1, I 2 : 58.8%, p<0.05). Further, sub-group analysis was performed by dividing the groups in only young/elderly participants. Young: The analysis for young participants performing gait with rhythmic auditory cueing revealed ( Supplementary Fig. 11) beneficial effects with large effect and substantial heterogeneity (g: 1.2, 95% C.I: 0.38 to 2.85, I 2 : 92%, p<0.01). Further, sub-group analysis with non-modulated rhythmic auditory cueing revealed ( Supplementary Fig. 12), under a single task condition, revealed a large effect size with substantial heterogeneity (Hedge's g: 0.81, 95% CI: -0.5 to 1.7, I 2 : 88%, p<0.01). Further, analysis with fast paced stimuli revealed small effect size with substantial heterogeneity (g: -0.01, 95% C.I: -0.4 to 0.4, I 2 : 92.5%, p<0.01). The heterogeneity as stated before could be attributed to differential rhythmic stimuli utilized by studies. Moreover, none of the studies analyzing a slow-paced stimulus evaluated stride length. Dual task performance was analyzed in only one included study. Therefore, no further analysis could be carried out to evaluate the effects of higher information processing constraints on stride length. Old: The analysis for old participants performing gait with rhythmic auditory cueing revealed ( Supplementary Fig. 13) beneficial effects with medium effect and substantial heterogeneity (g: 0.39, 95% C.I: -0.01 to 0.78, I 2 : 77%, p<0.01). Further, sub-group analysis with non-modulated rhythmic auditory cueing revealed ( Supplementary Fig. 14), under a single task condition, revealed a small effect size with negligible heterogeneity (Hedge's g: 0.22, 95% CI: -0.03 to 0.46, I 2 : 10.5%, p>0.05). Further, one study each analyzed the effects of fast, slow paced stimuli amongst elderly and further couldn't be included in sub-group analysis [67,101]. Dual task performance with auditory cueing in elderly participants was analyzed amongst two studies [67,94], a small effect size with negligible heterogeneity (g: -0.03, 95% C.I: -0.64 to 0.56, I 2 : 0%, p>0.05).

Cadence
The meta-analysis on healthy patients revealed (Fig. 6) a large effect size in positive domain with moderate heterogeneity (Hedge's g: 1.2, 95% CI: 0.51 to 1.8, I 2 : 41.9%, p<0.01). Further, sub-group analysis was performed by dividing the groups in only young/elderly participants.
Young: Further, sub-group analysis with nonmodulated rhythmic auditory cueing revealed ( Supplementary Fig. 15), under a single task condition, revealed a large effect size with substantial heterogeneity (Hedge's g: 1.76, 95% CI: -0.29 to 3.8, I 2 : 93.2%, p<0.01). Only one study performed [67], rhythmic auditory cueing with fast pace and no study analyzed the effects with slow paced stimulus. Therefore, no additional analysis was carried out. Dual task performance was analyzed in only one included study. Therefore, no further analysis could be carried out to evaluate the effects of higher information processing constraints on cadence.

DISCUSSION
The primary objective of this present systematic review and meta-analysis was to synthesize the current state of knowledge for effects that rhythmic auditory cueing might lay over aging gait. Out of thirty-four included studies, 88% studies reported beneficial effects of rhythmic auditory cueing on primary spatiotemporal gait parameters.
Typically, spatiotemporal parameters of gait worsen with age [19,106]. Callisaya, Beare, Phan, Blizzard, Thrift, Chen and Srikanth [107], studied age associated decline in brain structure with gait performance, and linked a reduction in gait velocity, stride length, cadence with white matter atrophy, lesions, hippocampal atrophy, and gray matter atrophy with cerebral infarcts, respectively [107,108]. Moreover, research suggests that degenerative changes in the fronto-striatal circuits might add increasing bi-directional stress on automated control for posture, gait and cognitive processing [109][110][111]. Possibly, explaining the loss of gait rhythmicity in elderly (see also, Nombela, et al. [56]). Likewise, increased energy expenditure [108], weak musculoskeletal structure associated variability in muscle contraction, and force production add towards the woes [112]. The current metaanalysis reported enhancements in gait velocity (g: 0.68), stride length (0.39) and cadence (0.78), post application of rhythmic auditory cueing in elderly population groups. Likewise, beneficial effects of rhythmic auditory cueing were also observed in gait amongst younger population groups.
Several mechanisms have been suggested to ascertain the beneficial effects of rhythmic auditory cueing. Rizzo, Raghavan, McCrery, Oh-Park and Verghese [113] for instance, speculated that auditory entrainment while performing gait might act as an efficient distractor. In addition, the auditory entrainment might also have aided in reducing the errors while executing the gait [114,115]. Possibly, by acting as an external guidance for "heel-contact" and "push-off" timings. Moreover, application of auditory entrainment is believed to allow enhancement in gait performance by bypassing or facilitating the degenerated basal ganglia-motor loop via alternative pathways [116][117][118]. Cunnington, Iansek, Bradshaw and Phillips [119] suggested that the external stimulation by entrainment might surpass deficient pallidal-cortical projections, and can directly serve an input supplementary motor area, thereby reducing the onset of motor deficit and aiding in performance. Moreover, the external cueing has shown to allow modulation of neuromagnetic β oscillations in auditory cortex, cerebellum, inferior frontal gyrus, somatosensory area and sensorimotor cortex [120], and reduce hemispheric asymmetry [121]. Neuroimaging studies reveal enhance activation in inferior colliculi [122], cerebellum, brainstem [117,123], sensorimotor cortex [124,125], further instigating cortico-cerebellar network re-organization [126]. Another crucial factor that considerably influences the aging gait is "change in tempo". Neurophysiological analysis suggests, increased neuronal activation in fronto-occipital networks [127], and excitability of the spinal motor neurons by reticulospinal pathways, with fast-paced entrainment. A paced-stimuli is thought to reduce the response time, limit the stagnating effects of constant entrainment over fractal scaling of stride times from healthy 1/f structure [128][129][130], and optimizing the velocity and acceleration profiles of joint motions by scaling movement time [59].
The present-meta-analysis also observed enhancements in the spatiotemporal parameters while performing dual-tasks, for both age groups. According to literature, dual-task performance predisposes to gait instability and falls by increasing cognitive motor interferences, across age groups [8,[131][132][133]. Interpretations from our results suggest that rhythmic auditory cueing counteracts cognitive constraints imposed by cognitively demanding dual-tasks such as carrying a tray and that this cueing might be useful in counteracting fall while carrying out activities of daily living [8]. Lohnes and Earhart [67], suggested that co-performance of dual-tasks with rhythmic auditory cueing might allow enhancements (or even stability) in performance, by possibly freeing up cognitive resources for dual-task performance. The authors also mentioned the influence of task complexity across age groups. Possibly, the freed up cognitive resources might not be sufficient especially in elderly to perform complex dual-tasks, such as coin transfer [134], and sentence reciting tasks [135]. This might possibly explain the reduced dual tasks costs on gait performance in young participants. In addition, the enhanced performance could also be attributed as to how the participants might perceive the auditory entrainment based on their cognitive capabilities. Wittwer, Webster and Hill [136], and Thaut, Miltner, Lange, Hurt and Hoemberg [137], suggested a strong relationship in between the cognitive capabilities and the ability to interpret and discern the structure of a beat. Thereby, suggesting a better rhythmic perception and interpretation by younger population groups as compared to their older counterparts.
Moreover, the progressive degradation of neuromuscular structures with aging has further been suggested to alleviate the threshold for action relevant acoustic input [138]. To counteract this deficiency use of ecologically valid acoustic feedback has been suggested [138]. The ecologically valid action related sounds might enhance saliency of sensory information concerning spatiotemporal information, thereby aiding in movement execution [100,[138][139][140][141]. This was also demonstrated by Dotov, et al. [100], here the authors demonstrated beneficial effects in parkinsonian and healthy gait parameters with biologically variable rhythmic auditory cueing as compared to isosynchronous cueing. Moreover, recent research has also revealed the possibilities of including emotional [113], motivational [68], and expressiveness [142], component in auditory entrainment to portray differential effects on gait parameters. Unfortunately, lack of pertinent, repeatable literature concerning the specific type of modified auditory feedback makes it difficult to interpret, as to which type of feedback might be most optimal, and for which age groups. We suggest future studies to replicate data concerning the use of ecological auditory entrainment across different age groups, to allow a reliable interpretation, which could then be included in gait rehabilitation protocols. Moreover, we also suggest future researchers to analyze the "entrainment effects" while multitasking in high-stress situations pertinent to modern day scenarios (for example, walking and texting, listening to music while crossing a traffic light).
This current meta-analysis also reported an increase in coefficient of stride-time and length variability in elderly participants with rhythmic auditory cueing. Based, on the published literature initial increase in variability during learning paradigm is efficient for improving gait performance [143]. Here, interpretations could possibly be drawn from "dynamic system theory" [144]. The theory suggests that a biological system might allow variability to identify and self-organize the most stable and viable outcome [144,145]. Thereby, interpretations could be made for regulating gait amongst young and elderly population groups to regulate gait when passing through fall-prone environments [41]. The present metaanalysis did not evaluate the the influence of gait training with rhythmic auditory cueing on ageing gait. Whereas, training regimes with auditory entrainment have demonstrated reduced variability in parkinsonism [101,146], and stroke [126]. We suggest future research to address this gap in the literature and evaluate the effects of long term training with rhythmic auditory cueing on aging gait.
Finally, we believe that the benefits of auditory entrainment might surpass that of co-treatment techniques (for instance, biofeedback, virtual reality, physiotherapy etc.) because of its economical nature, and high viability [77,78]. The rhythmic entrainment factor could be utilized with music in rehabilitation, day to day lives. This could allow benefits in both psycho-physiological domains [147][148][149][150][151]. For instance, improving stress, mediating arousal, emotions, internal motivation, memory, attention, executive functions [152], power [153], and endurance [154]. Moreover, it is important to consider that the retention of enhancements in gait parameters relies not only on the training received in the clinic but also depends largely on how much the patient follows the treatment protocol at home. Lim, et al. [13] for instance, reported enhancement in parkinsonian gait activity to 35 minutes per day (qualifying the 30 minutes criteria by WHO [155]). We believe that delivering this type of home-based intervention could possibly be beneficial for people lacking proper exposure to medical interventions in developing countries [156]. For instance, a booming number of smartphone devices in developing countries [157], can be used as a delivery tool while using a simple metronome app such as, Walkmate [129], or Listenmee [158], which with proper medical guidance might allow curbing the motor deficits associated with aging [159]. We also suggest the use of rhythmic auditory cueing as an adjunct to other rehabilitation strategies, for instance, dance, tai-chi, aerobics, as it might enhance the rehabilitation progress by focusing on both psychophysiological components.
To the best of our knowledge, this present review for the first time analyzed the effects of auditory entrainment on aging gait. The present findings are in agreement with systematic reviews and meta-analysis carried out to analyze auditory entrainment effect on stroke [66], cerebral palsy [160], and parkinsonism [57,161]. In conclusion, this review strongly suggests the incorporation of rhythmic auditory cueing for enhancing gait performance with aging gait. The results from the meta-analysis also direct towards the possible use of auditory entrainment to reduce the incidence of falls in high-stress situations.

Competing Financial Interests
No financial interests are declared.