The University of Manchester Research

The effect of rhythmic musical training on healthy olderadults' gait and cognitive functionDOI:10.1093/geront/gnt050

Document VersionAccepted author manuscript

Link to publication record in Manchester Research Explorer

Citation for published version (APA):Maclean, L. M., Brown, L. J. E., & Astell, A. J. (2014). The effect of rhythmic musical training on healthy olderadults' gait and cognitive function. Gerontologist, 54(4), 624-633. https://doi.org/10.1093/geront/gnt050

Published in:Gerontologist

Citing this paperPlease note that where the full-text provided on Manchester Research Explorer is the Author Accepted Manuscriptor Proof version this may differ from the final Published version. If citing, it is advised that you check and use thepublisher's definitive version.

General rightsCopyright and moral rights for the publications made accessible in the Research Explorer are retained by theauthors and/or other copyright owners and it is a condition of accessing publications that users recognise andabide by the legal requirements associated with these rights.

Takedown policyIf you believe that this document breaches copyright please refer to the University of Manchester’s TakedownProcedures [http://man.ac.uk/04Y6Bo] or contact uml.scholarlycommunications@manchester.ac.uk providingrelevant details, so we can investigate your claim.

Download date:02. Jun. 2020

The effect of rhythmic musical training on healthy older adults’ gait and cognitive

function

This is a pre-copyedited, author-produced PDF of an article accepted for publication in The Gerontologist

following peer review. The final published version of the article Maclean, Brown & Astell (2014) is

available online at: http://dx.doi.org/10.1093/geront/gnt050

Lin M. Maclean1, Laura J. E. Brown, PhD

2, and Arlene J. Astell, PhD

1

1. School of Psychology and Neuroscience, University of St. Andrews, St. Andrews,

Scotland

2. Department of Psychology and Research Institute for Health and Social Change,

Manchester Metropolitan University, Manchester

Corresponding author:

Linda Maclean

School of Psychology and Neuroscience

University of St. Andrews

St Mary’s Quad

South Street

St. Andrews, KY16 9JP

Phone : 0044 (0)1334 461990

Fax: 01334 463042

Email: lmm18@st-andrews.ac.uk

1

ABSTRACT

Purpose of the study: Older adults’ gait is disturbed when a demanding secondary cognitive

task is added. Gait training has been shown to improve older adults’ walking performance but

it is not clear how training affects their cognitive performance. This study examined the

impact on gait, in terms of cost or benefit to cognitive performance, of training healthy older

adults to walk to a rhythmic musical beat.

Design and Methods: In a mixed model design, forty-five healthy older adults over 65 years

of age (M = 71.7 years) were randomly assigned to three groups. One group received a

rhythmic musical training (MT) and their dual-task walking and cognitive performances were

compared with a group who had music playing in the background (MP) but no training, and a

third group who heard no music and received no training (NM). Outcomes in single (ST) and

dual-task (DT) conditions were step-time variability and velocity for gait and correct

cognitive responses for the cognitive task.

Results: The musical training group’s step-time variability improved in both the ST (P

< .05) and in the DT (P < .05) after training, without adversely affecting their cognitive

performance. No change was seen in the control groups.

Implications: Rhythmic musical training can improve gait steadiness in healthy older adults

with no negative impact on concurrent cognitive functioning. This could potentially enhance

‘postural reserve’ and reduce fall risk.

Key terms: attention; dual-task; musical-training; cognition

2

Introduction

The ability to carry out two tasks simultaneously is achieved by efficient allocation of

attention to both activities. Allocating attention is an executive function carried out within

working memory (Baddeley, 1986) the system that oversees aspects of higher-level cognitive

function (Coolidge & Wynn, 2005). The dual-task paradigm provides a means of examining

the allocation of attention when carrying out two tasks simultaneously, by comparing the

impact on performance relative to carrying out the two tasks singly (Baddeley & Hitch,

1974). There is almost always a ‘cost’, known as the dual task deficit (DTD) of dividing

attention between two tasks no matter how ‘simple’ they are (Smith & Kosslyn, 2007).

Walking is a well-practised motor-action that involves some element of attention, but with

increasing age, it becomes less automatic and more attention is required (Dubost et al., 2006;

Lindenberger & Baltes, 2000). This is even more pronounced when walking is combined

with a second activity such as talking. Decreased ability to allocate attention when carrying

out two tasks, such as walking and talking simultaneously may be a marker of cognitive

decline (Yogev-Seligmann, Rotem-Galili, Dickstein, Giladi & Hausdorff 2012a). It has been

suggested that healthy older adults unconsciously adopt a ‘posture first’ strategy when

simultaneously walking and carrying out a cognitively demanding task (Shumway-Cook,

Woollacott, Kerns & Baldwin, 1997). That is, they naturally default to attending more to their

gait than to the cognitive task, presumably to ensure their gait (Yogev-Seligmann et al.,

2010). As such, reduced ability to effectively allocate attention can be detrimental to posture

and balance, increasing the risk of falls (Siu, Choua, Mayrb, van Donkelaara & Woollacott,

2009).

When the cognitive load becomes very demanding, gait stability (which includes a measure

of step-to-step variability) suffers (Hausdorff, 2005). This is important because dual-task

situations, such as walking and talking, are common in everyday life. Finding ways to

3

improve older adults’ attention allocation could be beneficial both for ensuring gait stability

and enhancing ‘postural reserve’, which is an individual’s ability to respond effectively to a

postural threat (Yogev-Seligmann, Hausdorff & Giladi, 2012b).

Interventions to improve attention-allocation when walking and carrying out a cognitive task

have been developed for people with Parkinson’s disease (PD). Some have shown that

training older adults with PD to optimise their division of attention when walking and

carrying out another activity can benefit their gait (Yogev-Seligmann, Giladi, Brozgol &

Hausdorff, 2012c). Others demonstrate that rhythmic movement training increases gait

regularity and automaticity, which, in turn, is linked to increased gait safety plus a reduction

in fall risk (Bridenbaugh & Kressig, 2010). Auditory cues, sometimes embedded in music,

have been particularly successful at training older adults with PD to walk more steadily

(Rochester, Burn, Woods, Godwin & Nieuwboer, 2009; Satoh & Kuzuhara, 2008; Thaut,

McIntosh, Rice, Miller & Rathbun, 1996; Trombetti et al., 2011). Matching auditory cues to

their preferred walking pace optimises steady gait in people with PD (Arias & Cudeiro, 2008;

de Bruin et al., 2010).

Improving gait steadiness could also benefit healthy older adults who do not currently have a

gait or cognitive disorder. Training older adults to walk more steadily by making gait more

rhythmic and automatic so that it requires less mental effort could free attentional resources,

which potentially could be transferred to carrying out a secondary cognitive task (Luszcz,

2011; Moors & Houwer, 2006;Yogev-Seligmann, Hausdorff & Giladi, 2008 ).

4

Methods

Ethics

The study was approved by the University of St Andrews Teaching and Research Ethics

Committee (UTREC). Participants had the opportunity to raise questions about the study

before providing signed consent to participate.

Participants

Forty-five participants (28 women, 17 men), aged between 65-88 years (mean 71.7 years)

were recruited. Inclusion criteria were physically and cognitively healthy adults over 65

years, able to walk unassisted, living independently and speaking English as a first language.

They were assigned to one of three groups: the first 15 to a Musical Training group (MT; n =

15), the next 15 to a Music Playing group (MP; n = 15) and the final 15 to a No Music group

(NM; n = 15).

Design

The experiment used a mixed model design. The independent variables were time (pre- vs.

post-intervention training) and group (MT vs. MP vs. NM). The dependent variables for gait

were step-time variability and velocity in both single (ST) and dual task (DT) conditions. The

dependent variable for cognition was the number of Correct Cognitive Responses (CCR).

Materials

A demographic questionnaire was constructed to collect self-reported health, mood and

lifestyle measures. A battery of norm-referenced cognitive measures was assembled to

provide a baseline assessment of participants’ functioning, including memory and executive

functions (Table 1).

5

Insert Table 1 about here

Equipment

The ‘Bigfoot’ footswitches and connected software were developed in the School of

Psychology at the University of St Andrews. Bluetooth technology, housed in a PC ‘mouse’

in a box attached to the participant’s waist, was used to measure mean step-time, velocity,

number of steps and step-time variability via two footswitches which were fastened to their

heels with ‘Velcro’ strips and attached by wires to the box containing the ‘mouse’. The

‘Bigfoot’ equipment had previously been validated against video recording equipment and a

stopwatch to measure CV and velocity (Maclean & Astell, 2012).

The music chosen for training was the ‘Bluebell Polka’, which has a 2/4 rhythm (two strong

beats to each bar). The preferred walking pace of each of the participants in the Music

Training and the Music Playing Groups was matched to the music using the ‘Amazing Slow

Downer’ downloaded from www.ronimusic.com This software allowed the timing of the

selected music to be slowed down or quickened from the 100% pace set on the original

programme. The adjusted music was placed on a ‘loop’ and played continuously. The

researcher could then mute and unmute the music at will, depending on the experimental

condition.

Procedure

Participants were asked to attend wearing comfortable, flat shoes. Each participant first

completed the demographic questionnaire, cognitive test battery and BDI. Participants then

completed a short balance and mobility battery. They were then asked to walk along a 15m

indoor walkway twice at their self-selected pace. The walkway was a flat corridor marked off

6

at either end with tape indicating the 15m walk plus 1m at either end to allow for acceleration

and deceleration. The ST time and number of steps were averaged and used to adjust the pre-

selected music to the participant’s preferred walking pace.

The ST cognitive condition was a one-minute seated test in which participants performed a

Serial 7s test from a randomly generated three-digit number. They were allowed to practise

and only the second score was recorded. The ST test was counter-balanced between the start

and end of the experiment, for each group, to minimize possible practice effects during the

DT.

In the pre-intervention DT participants in all groups walked at their own pace whilst

subtracting 7s out loud (DT) from an initial three digit figure, which was randomly generated

for each DT condition. The DT walking conditions were carried out twice so that participants

could practice the secondary cognitive task, with data from the second task recorded. The

initial ST walking conditions were considered to be practice for the other pre and post–

intervention ST walking conditions. All of the cognitive tasks (both ST and DT, practice and

recorded) started with different 3-digit numbers.

Musical Training

The Musical Training Group was instructed to walk in time to the adjusted music.

Participants were not told that the music had been adjusted to suit their preferred walking

pace. Each participant walked up and down the 15m until they felt that they were walking

naturally in time to the music. If clarification was required, they were told to walk in time to

the music, until they were no longer thinking about the music or the walking. Participants in

the MT group walked an average of 4 times until they felt they were walking comfortably to

the music ‘without having to think about it’. The post-intervention conditions consisted of a

7

ST 15m walk with the adjusted music playing, followed by a DT 15m walk with the adjusted

music playing.

The Music Playing Group carried out all the same conditions as the Musical Training Group

without being trained to walk to the music. This group was told that there would be music

playing in the background, but were not instructed to walk to the music. The same

intervention (15 metres x 4 times) was used for this group as for the MT Group and, like the

MT group, these participants were not told that the music had been adjusted to suit their

preferred walking pace.

The No Music Group completed all the same conditions as the other two groups but there

was no music playing throughout the experiment.

Each group completed the walking conditions in the same order to ensure everyone

experienced the 4 x 15 metre walks at the same point in the experiment. The main gait

parameters measured over 15m were: mean step-time (ms) variability, expressed as

Coefficient of Variation (CV = M/SD) in ms and velocity expressed as m/s. The main

cognitive parameters for each DT were: participants’ correct responses per second (calculated

as the total number of responses divided by the time of each walk, multiplied by the

proportion of correct answers to the total number of answers, thereby accounting for number

of errors made) and number of steps taken per cognitive response, as a measure of the pace at

which participants simultaneously walked and counted.

Statistical analyses

Responses to the questionnaires and standardised cognitive tests were scored and compared

between groups using parametric analyses. Descriptive statistics indicated that the gait data

were non-parametric. The Kruskal-Wallis was used to test for group effect and the Mann-

Whitney (post-hoc) test was used to test for differences between pairs of conditions across the

8

groups. Wilcoxon signed-rank (post-hoc) test was used to compare results between pairs of

conditions within groups. The statistical significance for the experiment was set at 0.05.

Effect sizes were calculated and reported as r values. Using G*Power Version 3.1, we

calculated that 45 participants would be needed in this experiment, giving 0.8 power to detect

a large effect size of 0.5.

Results

The 45 older adults were healthy, with the majority reporting only one age-related health

condition plus very small numbers of hospital visits in the past twelve months alongside high

self-reported general satisfaction with their health (Table 2). Eighteen out of the forty-five

participants had never smoked, 29/45 reported a moderate level of alcohol intake (less than 7

units of alcohol per week) and 16/45 had a normal Body Mass Index (BMI: Table 2). Forty-

one out of forty-five, walked every day and 43/45 self-reported being cognitively and

physically active more than three times a week (Table 2). On the mobility measures, 12 out of

the 45 healthy older adults, 4 in the MT group, 3 in the MP group and 5 in the NM group,

chose not to attempt the heel-to-toe walk. This has previously been taken as an indicator of a

fear of falling (Nakamura, Holm &Wilson, 1999),but closer inspection of the falls history of

these 12 participants did not support this concern. Indeed, 9 of the 12 had had no falls in the

previous year, 1 had had one fall and 2 had more than one fall. Additionally, another 8 people

of the remaining 33 reported one or more falls in the past 12 months but they did not decline

to complete the heel-to-toe walk.

Insert Table 2 about here

9

The participants’ self-reported cognitive function was supported by their performance on the

battery of cognitive measures (Table 2). All 45 participants were in the normal range of

global cognitive function as measured by the MMSE and their pre-morbid IQs suggested they

were an above average sample. None was suffering from depression as measured by the BDI

(Table 2).

Cognition

Before training the three groups performed similarly at counting backwards in 7s both singly

and when walking and counting (Table 3). Although the intervention training did not improve

the MT group’s cognitive performance in the dual task, as it was expected to do, the MT

group’s DT cognitive scores did not decrease as a result of the musical training’s impact on

gait. The MP and NM groups also showed no change in their DT cognitive performance after

the intervention.

Gait

In the dual-task condition, all three groups showed a significant dual-task deficit (DTD) in

velocity in the pre-training conditions (Table 3; pre-training MT group P < .05, r = -.52, MP

group P < .001, r = -.62, NM group < .001, r = -.62. In the post-training conditions the MP

group P < .001, r = -.61 and the NM group < .001, r = -.59 had a DTD but the MT group’s

post-training DT speed (M = 1.08) was not significantly different from the ST post-training

performance (M = 1.09) P > 0.05. That is, after intervention training, the MP and the NM

groups’ dual-task ‘cost’ was that both groups walked more slowly when performing the gait

and cognitive tasks together, relative to when they walked without counting, but the MT

group who, after training, showed no DT deficit speed, relative to the same ST condition.

Gait stability was examined by looking at CV of step-time variability between groups. At

baseline, i.e. before training, the MT group’s CV (Mdn = .130) was significantly higher

10

(more unsteady) in both the ST (U = 53.5, P < .05, r = -.44) and DT (Mdn = .280: U = 49.5,

P < .05, r = -.48) conditions than the MP group (Mdn = .050) (Table 3). The MT group’s gait

was not significantly less steady than the NM group at baseline. After training, the step-time

variability of the Musical Training group (Mdn = 0.130) improved significantly in the ST

(Mdn = 0.060; T = 0, P < .05, r = -.62) whereas there was no significant change in CV in

either the MP or NM groups’ dual-task step-time variability compared to the ST (Table 3).

The CV of the MT group in the DT condition also improved significantly after training (T =

16.5, P < .05, r = - .41; Table 3), whereas there was no change in either of the other two

groups. The improvement in the MT group was such that after training, their gait became

steadier in the DT (Mdn = 0.07) than it had been in the ST at baseline (Mdn = 0.13), when the

walking task was performed alone. The significant improvement in DT CV was produced at

no ‘cost’ to DT cognition, i.e. there was no decline in performance on the secondary

cognitive task in the DT condition (Table 3).

Insert Table 3 about here

Attention Allocation

The magnitude of pre- to post-intervention change in step-time variability (CV) in the ST and

DT conditions was compared between groups (Table 3). This revealed a significant group

difference in the DT H (2) = 8.14, P < .05. Post hoc tests revealed that the MT group’s CV

improved significantly more from pre- to post-training DT performance than either the MP

group (U = 58, P < .05, r = -.34,) or the NM group (U = 55, P <.05, r = -.36; Figure 1).

Insert Figure 1 about here

11

To assess whether musical training freed up attention during the DT, we examined the

percentage improvement or decline associated with the MT group’s ST and DT pre- and post-

intervention training performances to provide a proxy measure of mental effort. The MT

group’s gait variability in the ST (pre- to post-training) improved by 38.9% after musical

training whereas it improved by 53.1% in the DT after musical training, with no concurrent

detrimental effect to the performance in the cognitive task (Figure 2). If the amount of effort

applied to gait is constant across the ST and DT, then the difference between the two

conditions, 14.2%, could be seen as a measure of the participants’ mental effort in the DT

that went on the secondary cognitive task (Figure 2).

Insert Figure 2 about here

The MT group’s pre-training gait in the DT was 43.7% less steady than in the ST (DTD).

After training, it was only 26.7% less steady in the DT than in the ST (DTD). The steadier

gait in the DT post-training suggests that 17% mental effort (the difference between ST-DT

pre-training and ST-DT post-training performances) was ‘released’ by the musical training

and re-directed back to gait. (Figure 2).

Discussion

As predicted, the gait of the musical training (MT group) became steadier after the

intervention training, as demonstrated by a significant improvement in CV between pre- and

post-training conditions in both the single and dual tasks and no DTD in velocity after

intervention training. In other words, gait velocity after training did not decline in the DT

relative to the ST. These two results for gait cannot be attributed simply to hearing music, as

there was no change in the gait of the MP group who had music playing in the background as

they walked and counted. The results also cannot be explained as a practice effect as there

12

was no improvement in the gait of the NM group who completed the walking and counting

conditions with no music playing. Moreover, the only change in gait between ST and DT, pre

to post-training, across the three groups, was the improvement in the MT group’s DT step-

time variability (CV), which became steadier in the DT than in the initial baseline ST. At

baseline the MT group’s CV gait was less steady in both the ST and DT pre-training stages

than the other two groups, reflecting the variable nature of gait. This is also seen in the non-

normal distribution of scores found in all of the gait conditions, which reflect the variability

of gait speed and steadiness between individuals (Hausdorff, 2005).

The results indicated that the MT group was a group of older community-dwellers, with no

known physical or cognitive impairments, randomly selected and allocated to experimental

groups and, as such, constitute a representative sample. Taken together, these results suggest

that musical intervention training improves gait stability, whilst having music playing

(controlling for the musical rhythm) and simply walking (controlling for the music) as

interventions have no effect on steadiness of gait. We anticipated that if musical training

improved gait stability by making it more automatic, this would free up controlled attentional

processing, which could in turn be transferred to performance of the more difficult cognitive

task (Schneider & Shiffrin, 1997). Between the pre- and post-training conditions, there was

no significant difference in the cognitive performance of the MT group, demonstrating

neither decline nor improvement as their gait became steadier.

Finding a way to quantifying the mental effort exerted on either the walking or the secondary

cognitive task is a major challenge (Yogev-Seligmann et al., 2012a). It may be possible to

examine this by measuring the magnitude of change between pairs of conditions. Looking at

the group that showed significant change after training (MT group), we can make two

observations about the magnitude of change between pairs of conditions. Firstly, the

difference between the ST pre- and ST post-training conditions suggests that the musical

13

training ‘saves’ 38.9% mental effort when walking. The size of improvement between the

DT pre and DT post-training conditions suggest that 53.1% of mental effort is saved by

musical training when walking and counting. The difference between the two ‘savings’

indicates that 14.2% of mental effort was used to count in the DT. Secondly, the

improvement between the ST pre-intervention (walking only) and the DT pre-intervention

(walking and counting) and ST post-intervention (walking to music) and DT post-

intervention (walking and counting to music) gait variability demonstrates that the MT group

had 17% extra mental effort to spare, due to the musical training, which could have been used

for the cognitive task but wasn’t. This suggests that the freed-up attention (17%) was not

automatically allocated to the cognitive task but possibly, unconsciously, re-directed to the

primary walking task.

There are two possible explanations for the Musical Training group’s correct cognitive

responses (CCR) remaining unaffected by the musical training. Firstly, during the DT, the

MT group participants, had reached their cognitive performance ceiling (Tehan & Mills,

2007) and no amount of freed-up attention could have increased the number of cognitive

responses produced by them. We suggest, therefore, that the extra attention freed by the

intervention training, for the MT group, was unconsciously allocated back to gait. An

alternative suggestion is that the MT group, being cognitively healthy, would naturally adopt

a posture first strategy in any situation where gait was threatened. In this case, it is possible

that, because the cognitive task was demanding, the threat to gait was maintained in the DT

and any extra attention, freed-up by the musical training and which might have produced an

improvement in CCRs, was, instead, unconsciously allocated to improve gait stability. This

added attention to gait could be what was observed in the MT group’s improved DT post-

intervention CV.

14

We also anticipated that the cognitive performance of the two groups who did not receive

music training would remain unchanged and this was the case. Since their gait did not

become any steadier either, this suggests that they were unaffected by the experimental

intervention.

Current understanding of gait suggests that it depends on the integrity of various aspects of

cognition, particularly the higher-level executive functions of which attention-allocation is

one (van Iersel, Kessels, Bloem, Verbeek & Olde-Rikkert, 2008). Working memory regulates

the attentional flow, which allows gait to be either consciously under control (‘effortful’) or

unconsciously automatic (‘effortless’) (Schneider & Shiffrin, 1997; Stuart-Hamilton, 2012).

The MT group’s improved gait performance suggests that rhythmic training may tap into the

ability of the working memory to flexibly process demands for attention and allocate it

according to a changing situation, for example when a secondary cognitive task becomes

more demanding (Baddeley, 1996; Yogev-Seligmann et al., 2008). Our findings support the

suggestion that walking while listening to music may render gait more ‘automatic’ for

healthy older adults who would usually unconsciously favour gait in a DT.

The performance outcomes from this group of physically and cognitively healthy older adults

have relevance both for other healthy older adults and those with impairment in gait and/or

cognition. In the first case, ‘postural reserve’ is an important factor affecting task

prioritisation during gait and cognition and which also decreases with age and disease.

Improving healthy older adults’ gait by making it more rhythmic through musical training,

whilst maintaining cognitive capability, could result in enhanced postural reserve. Secondly,

if musical training frees up attention, this could be beneficial to older adults with cognitive

impairment, who naturally adopt a ‘posture second’ strategy, whereby they allocate attention

equally to gait and cognition, thus increasing their risk of falling (Bloem, Grimbergen, van

Dijk & Munneke, 2006). Therefore, musical training to make gait more rhythmic could be a

15

preventative intervention for fall risk for older adults who are adversely affected by ‘walking

when talking’ (Lundin-Olsson, Nyberg & Gustafson, 1997).

This experiment was conducted in a controlled environment over a flat surface, with healthy

older adults, but it has the potential to be applicable to in a more ecologically valid and

complex environment, such as outdoors or in the home. Further research is required to

establish the practical application and therapeutic value to practitioners of these findings.

FUNDING

This work was supported by a doctoral studentship awarded to the first author as part of grant

number ES/G008779/1 from the Economic and Social Research Council held by the third

author.

16

REFERENCES

Arias, P. & Cudeiro J. (2008). Effects of rhythmic sensory stimulation (auditory, visual) on gait in

Parkinson’s disease patients. Experimental Brain Research, 186(4), 589-601. doi:

10.1007/s00221-007-1263-y

Baddeley, A. D. & Hitch, G. J.(eds) (1974). Working Memory (Vol. 8). New York: Academic Press.

doi:10.1007/s00221-007-1263-y.

Baddeley, A. D. (1986). Working Memory. Oxford: Oxford Scientific Publications.

Baddeley, A. (1996). Exploring the Central Executive. The Quarterly Journal of Experimental

Psychology, 49A(1), 5-28. doi: 10.1080/713755608.

Beauchet, O. (2006). Relationships between dual-task related changes in stride velocity and stride

time variability in healthy older adults. Human Movement Science, 25, 372–382.

Beck, A. T., Steer, R.A. & Brown, G. K. (1996). Manual for the Beck Depression Inventory-II. San

Antonio, TX: Psychological Corporation.

Bridenbaugh, S. A. & Kressig, R. W. (2010) Laboratory Review: The Role of Gait Analysis in

Seniors’ Mobility and Fall Prevention. Gerontology, 1-9.

Coolidge, F. L. & Wynn, T. (2005) Working memory, its executive functions and the emergence of

modern thinking. Cambridge Archaeological Journal,15(1), 5-26. doi:

10.1017/S0959774305000016.

de Bruin, N., Doan, J. B., Turnbull, G. Suchowersky, O., Bonfield, S., Hu, B. & Brown, L. (2010).

Walking with music is a safe and viable tool for gait training in Parkinson's disease: The

effect of a 13-week feasibility study on single and dual-task walking. SAGE-Hindawi Access

to Research Parkinson's Disease. doi: 10.4061/2010/483530.

Dubost, V., Kressig, R.W., Gonthier, R., Herrmann, F.R., Aminian, K., Najafi, B., Beauchet,O.

(2006). Relationships between dual-task related changes in stride velocity and stride time

17

variability in healthy older adults. Human Movement Science 25, 372–382. doi:

10.1016/j.humov.2006.03.004.

Folstein, M., Folstein, S. & McHugh, P. (1975) "Mini-mental state". A practical method for grading

the cognitive state of patients for the clinician. J Psychiatr Re,12, 189-198.

Hausdorff, J. Gait variability: methods, modeling and meaning. (2005). Journal of NeuroEngineering

and Rehabilitation, 2(1), 19. doi:10.1186/1743-0003-2-19.

Hausdorff, M. (2010). How does explicit prioritization alter walking during dual-task performance?

Effects of age and sex on gait speed and variability. Pysical Therapy, 90(2),177-186. doi:

10.2522/ptj.20090043.

Lindenberger, U., Marsiske, M. & Baltes P. (2000) Memorizing while walking: increase in dual-task

costs from young adulthood to old age. Psychology and Aging, 15, 417 - 436.

doi:10.1037/0882-7974.18.2.181.

Lundin-Olsson, L., Nyberg, L., Gustafson, Y. (1997) “Stops Talking when talking” as a predictor of

falls in elderly people. The Lancet, 349, 617. doi: 10.1016/S0140-6736(97)24009-2.

Luszcz, M. (2011). Executive Function and Cognitive Aging. In: Schaie, K. W. & Willis, S. L.

(eds). Handbook of the Psychology of Aging, 7th Ed. London: Adademic Press, 59-72.

Maclean, L. M. & Astell, A. J. (2012) Are you paying attention? What walking and counting can tell

us about ageing. ISPGR/Gait & Mental Function, 1st Joint World Congress, , Trondheim,

0.6.6.

http://ispgr.org/fileadmin/templates/_img/norway2012/ISPGR_full_abstract_listing_SOP_20

12_07_31_web.pdf

Moors, A. & De Houwer, J. (2006) Automaticity: theoretical and conceptual analysis. Psychological

Bulletin,132, 297-326.

Nakamura, D. M., Holm, M. B., & Wilson, A. (1999). Measures of Balance and Fear of

18

Falling in the Elderly. Physical & Occupational Therapy in Geriatrics, 15(4), 17-32.

doi:10.1080/J148V15n04_02.

Nelson, H. E. & Willison, J.R. (1991). The Revised National Adult Reading Test - Test Manual, 2nd

Ed. Windsor: NFER - Nelson;

Reitan, R.M. & Wolfson, D. (1993) The Halstead–Reitan Neuropsychological Test Battery: Theory

and clinical interpretation; 2nd Ed. Tucson, AZ: Neuropsychology Press.

Rochester, L., Burn, D. J., Woods, G., Godwin, J. & Nieuwboer, A. (2009) Does auditory rhythmical

cueing improve gait in people with Parkinson's disease and cognitive impairment? A

Feasibility study. Movement Disorders, 24(6), 839-845. doi:10.1002/mds.22400.

Satoh, M. & Kuzuhara, S. (2008). Training in Mental Singing while Walking Improves Gait

Disturbance in Parkinson’s Disease Patients. European Neurology. 60(5), 237-243.

Schneider, W. & Shiffrin, R. M. (1997) Controlled and automatic human information processing: II.

Perceptual learning, automatic attending, and a general theory. Psychological Review, 84,

127-190.

Shumway-Cook, A., Woollacott, M., Kerns, K. A. & Baldwin, M. (1997) The Effects of Two Types

of Cognitive Tasks on Postural Stability in Older Adults With and Without a History of Falls.

The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 52A(4),

M232-M240. doi:10.1093/gerona/52A.4.M232

Siu, K., Choua, L., Mayrb, U. van Donkelaara, P. & Woollacott, M. H. (2009). Attentional

mechanisms contributing to balance constraints during gait: during gait: The effects of

balance impairments. Brain Research, 1248, 59-67. doi: 10.1016/j.brainres.2008.10.078.

Smith, A. (1982) Symbol Digits Modalities Test. Los Angeles Western Psychological Services.

Smith, E. E. & Kosslyn, S. M. (2007). Cognitive Psychology Mind and Brain . New Jersey: Pearson

Education.

Spreen, O. & Strauss E. (1998). A Compendium of Neuropsychological Tests, 2nd Ed. Oxford:

Oxford University Press.

19

Stuart-Hamilton, I. (2012). The Psychology of Ageing, 5th Ed. London: Jessica Kingsley.

Tehan, G. & Mills, K. (2007). Working memory and short-term memory storage:what does

backward recall tell us? In: Osaka, N., Logie, R. H. & D'Esposito, M., (eds). Oxford: Oxford

University Press, 153-164.

Thaut, M. H., McIntosh, G. C., Rice, R. R., Brault, J. M., Miller, R. A. & Rathbun, J. (1996).

Rhythmic auditory stimulation in gait training for Parkinson's disease patients. Movement

Disorders,11(2),193-200. doi:10.1002/mds.870110213.

Thorndike, R. L., Hagen, E.P. & Sattle, J.M. (1987). The Stanford-Binet Intelligence Scale;. 4th

ed.

Chicago Riverside.

Trombetti, A., Hars, M., Herrmann, F. R., Kressig, R. W., Ferrari, S. & Rizzoli, R. (2011). Effect of

Music-Based Multitask Training on Gait, Balance, and Fall Risk in Elderly People: A

Randomized Controlled Trial. Arch Intern Med, 171(6),525-53

doi:0.1001/archinternmed.2010.446.

van Iersel, M. B., Kessels, R. P. C., Bloem, B. R. Verbeek, A. L. Olde, R. & Marcel G. M. (2008)

Executive Functions Are Associated With Gait and Balance in Community-Living Elderly

People. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences,

63(12), 1344-1349.

Wechsler, D. (1987). Wechsler Memory Scale - Revised (WMS-R. San Antonio, TX: Psychological

Corporation.

Yogev-Seligmann, G., Hausdorff, J. & Giladi, N. (2008) The role of executive function and attention

in gait. Movement Disorders, 23,329-342. doi:10.1002/mds.21720.

Yogev-Seligmann, G., Rotem-Galili, Y., Mirelman, A., Doclsteom, R., Giladi, N. & Yogev-

Seligmann, G., Rotem-Galili, Y., Dickstein, R. Giladi, N. & Hausdorff, M. (2012a). Effects

of explicit prioritization on dual task walking in patients with Parkinson's disease. Gait &

Posture, 35(4), 641-646. doi:10.1016/j.gaitpost.2011.12.016.

20

Yogev-Seligmann, G., Hausdorff, J. M. & Giladi, N. (2012b). Do we always prioritize balance when

walking? Towards an integrated model of task prioritization. Movement Disorders, 1-6.

doi:10.1002/mds.24963.

Yogev-Seligmann, G., Giladi, N., Brozgol, M. A. (2012c) Training Program to Improve Gait While

Dual Tasking in Patients With Parkinson's Disease: A Pilot Study. Archives of Physical

Medicine and Rehabilitation, 2,1-6. doi: 10.1016/j.apmr.2011.06.005.

21

Table 1

Description of Baseline Cognitive and Mood Tests

Cognitive Test Description

Mini-Mental State Examination (MMSE)

(Folstein, Folstein & McHugh, 1975)

The MMSE provides a brief measure of

global cognitive function that is scored out of

30 with norm-referenced adjustments for

participants over 80 years of age (Spreen &

Strauss, 1998).

Trail Making Test A and B (TMT A & TMT

B: AITB, 1944 (Reitan & Wolfson, 1993)

TMT A and B provide a measure of

executive functions assessed through speed

in seconds to complete both parts (Delta

TMT – B minus A).

Symbol Digits Modalities Test (SDMT)

(Smith, 1982)

SDMT provides a measure of processing

speed where participants match as many

numbers to symbols as possible in 90

seconds.

Digit Span Forwards and Backwards (DS/F

& DS/B) (Thorndike, Hagen & Sattle, 1987)

These tasks provide a measure of Working

Memory by repeating short strings of

numbers both forwards and backwards. The

score is based on the number of items

correctly repeated

22

Immediate and Delayed Story Recall

(Wechsler, 1987)

Story Recall provides a measure of episodic

memory by asking participants to recall

immediately and after a 30-mintue delay

story as much detail as possible from a story

read aloud to them. Scoring comprises total

number of ideas correctly recalled

immediately out of 25 and the percentage of

ideas recalled after the delay.

National Adult Reading Test - 2nd

Edition

(Nelson & Willison, 1991)

The NART provides a measure of estimated

pre-morbid intelligence through reading a list

of progressively difficult and phonetically

irregular English words. The total number of

errors is converted to a Full Scale IQ

equivalent.

Beck Depression Index (BDI) (Beck, Steer &

Brown, 1996)

The BDI screens for the presence of clinical

depression, where scores above 13 indicate

clinically significant problems.

23

Table 2

Participant Demographic Characteristics and Physical and Cognitive Functions at Baseline

Characteristics Music

Training

(MT)

Group #

Music Playing

(MP)

Group #

No Music

(NM) Group

#

Number of Participants

Gender (male/female)

Age (years)

Self-reported chronic medical

conditions

Self-reported nights in hospital in

previous year

Self-reported falls in previous year

Never Smoked

Less than 7 units of alcohol a week

BMI indicates overweight or obese

Walking every day

Cognitive activities at least 3

times/week

Exercise at least 3 times a week

Satisfied with health status

Number of complete stand-up/sit -

downs

Time taken in Heel-To-Toe Walk

15

4/11

73.2 (±5.36)

1.67

1.13

1.33

3/15

7/15

9/15

14/15

14/15

15/15

14/15

9.35 (± 2.12)*

42.32 (± 9.87)

15

6/9

69.1 (±3.37)

1.53

1.07

1.2

6/15

6/15

11/15

13/15

15/15

13/15

13/15

12.17 (±2.89)

48.84 (± 19.48)

15

7/8

72.9 (±6.490

2

1.07

0.53

9/15

11/15

9/15

14/15

14/14

15/15

10/15

11.83 (±4.02)

55.45(±18.61)

24

Number who completed Heel-To-Toe

Walk

Balance on one leg time (seconds)

Mini Mental State Examination

Beck’s Depression Inventory

Estimated Pre-Morbid IQ

Digit Score/Forwards

Digit Score/Backwards

Delta/Trail Making Test

Symbol Digit Modalities Test

Immediate Recall (Story)

Delayed Recall (% of Immediate

Recall)

10/14

15.54 (±10.5)

28.4 (±1.18)

1.2 (±1.38)

119.6 (±4.32)

10.6 (±2.09)

6.87 (±1.8)

50.5 (±28.3)

39.9 (±7.2)

11.1 (±3.8)

94.3 (±14.6)

12/15

20.99 (±9.9)

29.1 (±1.28)

1.33 (±1.98)

117.33 (±5.52)

10.4 (±2.7)

8.2 (±2.4)

48.1 (±25.4)

43.6 (±10.8)

11.4 (±2.92)

102.8 (±13.9)

10/15

19.29(±11.06)

28.7 (± 1.83)

0.47 (±0.74)

117.5 (±7.72)

10.6 (±2.6)

8.6 (±2.9)

34.1 (±19.3)

47.0 (±9.3)

9.9 (±5.3)

83.9 ±26.6)**

#Values are mean (± Standard Deviation)

* Significant difference between the MT and the MP group, p < .05

**Significant difference between NM and MP groups, p < .05

25

Table 3.

Group Performance in Gait and Cognition Tasks – Single and Dual Task, Pre and Post-Intervention Training

Music Training Group (n = 15)

Median (range)

Music Playing Group (n = 15)

Median (range)

No Music Group (n = 15)

Median (range)

Single Task Dual Task Single Task Dual Task Single Task Dual Task

Parameter

Pre-

Trg

Post-

Trg

Pre-

Trg

Post-

Trg

Pre-Trg

Post-

Trg

Pre-Trg

Post-

Trg

Pre-Trg

Post-

Trg

Pre-Trg

Post-

Trg

Velocity 1.22

(.77)

1.09

(.87)

1.06‡

(1.15)

1.08

(.81)

1.38

(.59)

1.36

(.53)

1.15‡‡

(.83)

1.16‡‡

(.85)

1.25

(.88)

1.31

(.75)

0.986‡‡

(.97)

1.12‡‡

(.98)

Step-time

variability

(CV)

0.130†

(.62)

0.060*

(.61)

0.282§

(.99)

0.070*

(.45)

0.050

(.31)

0.060

(.26)

0.070

(.27)

0.080

(.43)

0.060

(.43)

0.060

(.42)

0.060

(.90)

0.170

(.69)

Correct

Cognitive

Responses

0.324

(1.15)

0.352

(1.15)

0.323

(1.28)

0.415

(.74)

0.399

(.77)

0.343

(.65)

0.361

(.56)

0.348

(.65)

0.296

(.60)

‡ Significant change from ST to DT, P < .05

‡‡ Significant changes from ST to DT, P < .001

† Significant difference between MT’s ST pre-training CV and MP group’s ST pre-training CV, P < .05

§ Significant difference between MT’s DT pre-training CV and MP and NP groups’ DT pre-training CV, P < .01

* Significant change from ST pre to post-training, P < .05

* * Significant change from DT pre to post-training, P < .05

26

Figure 1. Changes across groups between median DT pre and post-training step-time

variability

*Change significant at P<.050

ms = milliseconds

CV = Coefficient of Variation (Standard Deviation/Mean)

*

*

27

Figure 2. The musical training group’s median step-time variability in ST and DT, pre and

post-training walking conditions.

CV = Coefficient of Variation (Standard Deviation/Mean)

ms = milliseconds

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

Single Task Pre-training

Single Task Post-training

Dual Task Pre-training

Dual Task Post-training

Med

ian

Ste

p-t

ime

va

ria

bil

ity

(CV

) in

ms