To cite: van Breda E,Verwulgen S, Saeys W, et al.Vibrotactile feedback as atool to improve motorlearning and sportsperformance: a systematicreview. BMJ Open SportExerc Med 2017;3:e000216.doi:10.1136/bmjsem-2016-000216

EB and SV contributedequally.

Accepted 3 May 2017

1Faculty of Medicine andHealth Sciences, Departmentof Rehabilitation Sciencesand Physiotherapy(MOVANT), University ofAntwerp, Antwerp, Belgium2Product Development,Faculty of Design Sciences,University of Antwerp,Antwerp, Belgium

Correspondence to

Dr Eric van Breda; eric.vanbreda@uantwerpen.be

Vibrotactile feedback as a tool toimprove motor learning and sportsperformance: a systematic review

Eric van Breda,1 Stijn Verwulgen,2 Wim Saeys,1 Katja Wuyts,1 Thomas Peeters,2

Steven Truijen1

ABSTRACTBackground Evidence concerning the use ofvibrotactile feedback for acquiring and learning newmotor skills is limited. Although various concepts andapplications for tactile feedback have been proposed,little is known about the suitability of this feedbackmechanism in sports training.Aim The goal of this systematic review was to gatherknowledge on the efficacy of the use of vibrotactilefeedback in improving sports performance skills.Design Systematic review.Methods Comprehensively searched databases were:PubMed, Cochrane and Web of Science. Studiesinvestigating the effects of using vibrotactile feedbackin sports training in healthy subjects were included inthis review.Results No consensus was found regarding thepositive effectiveness on performance using vibrotactilefeedback in a sports context. No evidence was foundthat the addition of tactile feedback is effective foracquiring new motor skills. None of the studies show asignificant learning effect.

INTRODUCTIONWhen an athlete performs a designatedmovement (motor task), any attempt isneurophysiological evaluated by scanningthe information originating from intrinsicfeedback (IFB) systems such as visual, audi-tory, proprioceptive and/or tactilefeedback (TFB). Those feedback(FB) systems analyse and adjust extremityposition either in real time (concurrent FB)or store the information during post-taskevaluation (terminal FB).1 Thus IFB leadsto the information regarding the outcomeof the motor task (knowledge of results) andprovides information of the entire processof a motor task (knowledge of perfor-mance).1 2 The only IFB system that is indirect contact with the environment istactile sensory system, known as TFB. TFBallows us to interact with the environmentthrough specific sensors in the skin. Theysense vibration, pressure, touch, textureand so on. TFB is complex and is primarily

composed of a skin sensation, enablingsubjects to localise and recognise individualtactile cues and collaborates closely with theproprioceptive and kinaesthetic system,that is, body part position recognition.3

Interestingly, TFB is the only FB systemthat is bidirectional, that is, it interacts withthe environment and simultaneously sensesand processes such interactions.4–6

Although TFB is essential for movement-controlled actions, it can also be adisturbing factor by creating a kind ofsensory overload mainly because it can alsobe handled consciously.7

In contrast to IFB, augmented feedback(AFB) serves as an add-on to the task-intrinsic information and is exclusivelydelivered by external cues.2 To date, it isimpossible to consider modern sportstraining, but also rehabilitation of certainpatient groups, without the use of AFB.AFB comes from external sources,for example, trainers/coaches, electronicdevices (displays, sounds), spectators and

What is already known?

" Motor tasks are controlled by intrinsic feedbackmechanisms.

" Augmented feedback (verbal instructions) inlearning new motor tasks is important in sports.

What are the new findings?

" The use of vibrotactile feedback in severalmedical and non-medical areas has well beenestablished.

" Vibrotactile feedback as feedback tool in sportshas not yet been supported by scientificevidence.

" There lies a huge potential for sports science tostudy feedback mechanisms by using vibrotactilefeedback as intervention tool.

van Breda E, et al. BMJ Open Sport Exerc Med 2017;3:e000216. doi:10.1136/bmjsem-2016-000216 1

Open Access Review

recently also haptic systems (robots, vibrotactile(VT) actuators), or a combination thereof.3 AFB hasalso been shown to play an important role in learningnew motor tasks by enhancing the effect of IFB.5 8 Thiscan be provided verbally (a coach provides detailedspoken instructions), visually (a coach demonstrates thespecific task) or haptic (a coach guides part of thetrajectory of the movement by touching the subject atkey parts of the motor task) cues. Mostly, however,AFB is offered after completion of a motor task, asterminal FB. Although it has been shown that terminalFB can play an important role in the learning processof motor tasks, the major drawback of this type of FB isthat it only provides information on the outcome ofthe motor task but not real time on the process as inconcurrent FB.5 Nowadays, technological progressallows us to provide information to an athlete inreal time. One of such methods is vibrotactile feedback(VTFB). During VTFB, a signal is generated by anactuator that applies vibrational stimuli to the skinguided by information on position of an extremity.9

VTFB is a form of concurrent FB and information canbe delivered by varying temporal parameters infrequency, amplitude, waveform and/or duration.10

The tight coupling between motor task and theresulting VTFB stimuli requires a minimal latency.11

Instant and real-time augmented VTFB makes itfeasible to provide spatial proprioceptive informationduring a complex motor task.12 In this sense, VTFBcan thus be used to guide a motor task and provideinstant FB. VTFB is, from a neurophysiological pointof view, an interesting source of FB, since it can bypassthe athletes’ visual FB system. The relatively small, low-weight and wireless potential makes it an excellentFB tool.3

Prior to the use of VTFB, consistent metaphors needto be learnt. Amemiya et al suggested that wearableVTFB devices should create realistic sensations.13 Inthat light, Spelmezan and Ziefle distinguished twotypes of encoding metaphors: a push and a pull meta-phor.14 The pull metaphor means that the extremity ofa subject reacts to a signal by moving towards the stim-ulus, whereas the push metaphor means that it pullsaway from the stimulus.The major aim of this present systematic review is to

evaluate the definitions, methodologies and use ofFB systems for learning motor tasks in sports. To thebest of our knowledge, this is the first systematic reviewon VTFB systems within sports.

METHODSDesignThis systematic review was conducted in accordancewith the PRISMA (Preferred Reporting Items forSystematic Reviews and Meta-Analyses) guidelines,which is an updated statement addressing the concep-tual and methodological issues of the originalQUOROM (Quality of Reporting of Meta-Analyses) statement.15

Search strategyPICO(S) components were used to establish theresearch question. This PICO(S), along with the corre-sponding eligibility criteria, is shown in table 1.To complete this review, three electronic databases

were used. PubMed (http://www.ncbi.nlm.nih.gov/pubmed/), Web of Science (WOS; https://webofknowl-edge.com/) and the Cochrane database (http:// www.cochranelibrary.com) were searched for peer-reviewedarticles.

Table 1 PICOS and eligibility criteria

Inclusion criteria Exclusion criteria

P Adults

Adolescents

Athletes or healthy subjects who practise sports regularly

Children (<10 years)

Elderly (>65 years)

Non-athletes

Unhealthy subjects

Any pathology related to movements

Animals

I VTFB —

C Non-VTFB —

O Heart rate

Joint angles

Motor patterns

Movement variability

S Randomised clinical trials, randomised controlled trials, clinical

trial, case reports, cohort study

Non-English, Dutch, French or German,

systematic review, meta-analysis

VTFB, vibrotactile feedback.

2 van Breda E, et al. BMJ Open Sport Exerc Med 2017;3:e000216. doi:10.1136/bmjsem-2016-000216

Open Access

Table

2Databasesearchstrategy

Database

Searchquery

October2015

March

2017

PubMed

(‘Feedback,Sensory’[Mesh]OR‘sensory

feedback’[AllFields]OR‘Touch’[Mesh]OR‘hapticfeedback’[AllFields]OR‘VT

feedback’[AllFields]OR‘electro-opticalfeedback’[AllFields])AND

((‘Athletes’[Mesh])OR‘athletes’[AllFields])ORbobsled[All

Fields]OR‘w

intersports’[AllFields]OR‘sports’[MeSH

Term

s]OR‘sport’[AllFields]OR‘rowing’[AllFields])AND(‘Learning

Curve’[Mesh]OR‘learningcurve’[AllFields]ORperform

ance[AllFields]OR‘ReactionTim

e’[Mesh]OR‘reactiontime’[AllFields]

ORaerodynamics[AllFields]ORsynchronization[AllFields]OR‘posture’[MeSH

Term

s]OR‘posture’[AllFields]OR‘m

otor

learning’[AllFields]OR‘SpatialLearning’[Mesh]OR‘SpatialLearning’[AllFields]OR‘Practice(Psychology)‘[M

esh]OR‘Practice

(Psychology)‘[AllFields]OR‘M

otorSkills’[Mesh]OR‘M

otorSkills’[AllFields]OR‘AthleticPerform

ance’[Mesh]OR‘Athletic

Perform

ance’[AllFields]OR‘PsychomotorPerform

ance’[Mesh]OR‘PsychomotorPerform

ance’[AllFields]OR‘Touch

Perception’[Mesh:noexp]OR‘TouchPerception’[AllFields]OR‘Sensory

Thresholds’[Mesh]OR‘Sensory

Thresholds’[AllFields]

OR‘DifferentialThreshold’[Mesh]OR‘DifferentialThreshold’[AllFields]OR‘latency’[AllFields])

211

214

Cochrane

(‘Sensory

feedback’OR‘Touch’OR‘hapticfeedback’OR‘VTfeedback’OR‘electro-opticalfeedback’)AND

(‘athletes’OR

‘bobsled’OR‘w

intersports’OR‘sports’OR‘rowing’)AND(‘posture’OR‘learningcurve’OR‘perform

ance’OR‘reactiontime’

OR‘aerodynamics’OR‘synchronization’OR‘m

otorlearning’OR‘SpatialLearning’OR‘Practice(Psychology)’OR‘M

otorSkills’

OR‘AthleticPerform

ance’OR‘PsychomotorPerform

ance’OR‘TouchPerception’OR‘Sensory

Thresholds’OR‘Differential

Threshold’)

77

Webof

Science

TS=((‘sensory

feedback’OR‘Touch’OR‘hapticfeedback’OR‘VTfeedback’OR‘electro-opticalfeedback’)AND(‘athletes’OR

‘bobsled’OR‘w

intersports’OR‘sports’OR‘rowing’)AND(‘learningcurve’OR‘perform

ance’OR‘reactiontime’OR

‘aerodynamics’OR‘synchronization’OR‘posture’OR‘m

otorlearning’OR‘SpatialLearning’OR‘Practice(Psychology)’OR

‘MotorSkills’OR‘AthleticPerform

ance’OR‘PsychomotorPerform

ance’OR‘TouchPerception’OR‘Sensory

Thresholds’OR

‘DifferentialThreshold’OR‘latency’))

74

75

van Breda E, et al. BMJ Open Sport Exerc Med 2017;3:e000216. doi:10.1136/bmjsem-2016-000216 3

Open Access

The Cochrane was used to find relevant clinical trialswithout date restrictions as search option. Using theadvanced search option, a selection was made to findclinical trials only. Conference papers were excluded.FB conditions where VTFB is used to only roboticallyguide someone, that is, the subject playing a passiverole, were also excluded from this review, along withFB conditions where VTFB is used as sensorysubstitution.In addition, we manually searched for relevant

articles based on article citations and reference lists inthe PubMed and WOS databases. The articles foundwere extracted and organised in EndNote Online.Duplicates were removed manually. The first searchwas performed in October 2015 and the date of thelast search was on 30 March 2017.The search strategy is shown in table 2.

Study selectionAfter implementation of the research strategy in alldatabases and the execution of the manual search,articles were extracted and organised in EndNoteOnline. Duplicates were deleted by hand. Theremaining articles were transferred to an Exceltemplate published by the KCE-Belgian Health CareKnowledge Centre (http://kce.fgov.be/). Two hundredand ninety-six (296) articles from the three used data-bases, 321 studies selected for screening and 25records from other sources (30 March 2017) weresequentially included or excluded according to theeligibility criteria. Two reviewers (KW and TP)performed eligibility assessment independently in anunblinded standardised manner.Thus, the inclusion criteria were:

1. athletes or healthy subjects who practise sportsregularly;

2. VTFB or haptic FB;3. heart rate (HR), joint angles, motor patterns, move-

ment variability;4. randomised clinical trials, randomised

controlled trials, clinical trial, case reports, cohortstudy, and so on;

5. full text written in English, Dutch, German andFrench;

6. study subjects are adolescents or adults.

Exclusion criteria were:

1. non-healthy or pathological population and animals;2. children (<10 years of age) and elderly (>65+);3. other languages than those mentioned above;4. meta-analysis, systematic reviews.

Experiments by Pressley and Ghatala indicated thatchildren younger than 10 years of age are less accu-rately aware of FB results. Consequently, this set ourlower age limit in our search strategy.16 The same goesfor the upper limit in adults as has been shown byRodrigue et al.17

The systematic screening of the articles was carriedout in two different phases. First, all search resultswere screened based on title and abstract in a prelimi-nary screening. If citations were considered potentiallyeligible and relevant, and if both reviewers (KW andTP) independently had agreed, full-text articles wereobtained. In the second phase, full-text articles wereevaluated. Detailed evaluation of the full-text publica-tion was necessary for 47 articles. Disagreementsbetween reviewers were resolved by consensus. If noconsensus could be reached, a third reviewer (SV)would decide.

Data extractionTwo reviewers (KW and TP) independently performeddata extraction by selecting relevant data and inte-grating it into two databases. The two databases werecompared and integrated into a final extraction table.Again, disagreements were resolved through discussionbetween KW and TP, and in case of disagreement athird author (SV) was asked to make a final judgement.None of the authors were contacted for further infor-mation when information was missing or unclearlyreported. The reason for the latter was that non-peer-reviewed unpublished information would, amongothers, hamper the systematic review process andcould affect the quality of this paper as such.

Table 3 EBRO classification of study results and recommendations: classification of the study results according to the

level of evidence

A1 Meta-analyses (systematic reviews) which include at least some randomised clinical trials at quality level A2 that

show consistent results between studies

A2 Randomised clinical trials of a good methodological quality (randomised double-blind controlled studies) with

sufficient power and consistency

B Randomised clinical trials of a moderate methodological quality or with insufficient power, or other non-randomised,

cohort or patient–control group study designs that involve intergroup comparisons

C Patient series

D Expert opinion

EBRO, Evidence-Based Richtlijn(guideline)Ontwikkeling(development).

4 van Breda E, et al. BMJ Open Sport Exerc Med 2017;3:e000216. doi:10.1136/bmjsem-2016-000216

Open Access

Items extracted from the studies were study design,age and number of subjects included, study characteris-tics (including branch of sports or movement studied,setting and FB conditions), VTFB conditions (includingnumber of vibrotactors, vibrotactor placement,vibrating frequency, stimuli duration, encoding signalsand possible bandwidth used), type of outcome meas-ures, and follow-up or retention and finally mainresults.

Risk of bias assessmentTwo independent reviewers (KW and TP) independentlyassessed all included articles.Suitability of randomisation and concealment of alloca-tion, blinding of patients, healthcare providers, datacollectors, and outcome assessors, and extent of loss tofollow-up were determined by the two above-mentioned reviewers using the PEDro (PhysiotherapyEvidence Database) scale. According to de Morton,18

the PEDro scale is a valid tool to measure the method-ological quality of clinical trials.The risk of bias was assessed by the same authors

(KW and TP) who performed the systematic search ofliterature. The two authors performed the assessmentof methodological quality independently. Findings

were compared for agreement but in case of disagree-ment, a third judgement was in all cases obtained byone of the other coauthors (SV). Agreement betweenthe two reviewers concerning the scoring of methodo-logical quality was determined by the Cohen’s kappa.Cohen’s kappa coefficient is a statistic measure forinter-rater agreement for qualitative items. In all cases,disagreements were resolved by discussion.Scores were assigned based on meeting up with the

criteria. A maximum possible score of 10 (low risk ofbias) could be reached. Scores 6–10 were regarded ashigh-quality studies, 4–5 as fair quality and 3 or loweras poor quality. Studies with a risk of bias assessmentscore of 3 or less were considered poor and theircontribution to results was weighted as half.Evidence was graded according to EBRO (Evidence-

Based Richtlijn[guideline]Ontwikkeling[development])recommendations (table 3).19 EBRO is an initiative ofthe Dutch Cochrane Centre and the Dutch Institute forHealthcare Improvement (CBO, http://www.cbo.nl), amember of the Guidelines International Network.

Data analysisSince the average quality of the studies was poor, aquantitative or meta-analysis was not feasible. We tried

Figure 1 Flow chart of the search strategy and study selection process. WOS, Web of Science.

van Breda E, et al. BMJ Open Sport Exerc Med 2017;3:e000216. doi:10.1136/bmjsem-2016-000216 5

Open Access

Table

4Data

extractiontable

Study

Design

Study

population

(n)

Mean

age

(years)

Movement

orbranchof

sports

Setting

Feedback

conditions

Training

(Noof

days)

Retention

Outcome

measures

Main

results

Eid

etal21

Clinicaltrial

10

—Treadmill

running

Lab

VTFB*

Nike+

SportBand

(VFB)

Nodevice

1None

HR

VTFB="efficient

regulatingHR(#

fluctuations)

VTFB=biggersteps,

breathing"efficient

FasterHRrecovery

Lieberm

anand

Breazeal22

Clinicaltrial

40

—5dfarm

motions

Lab

VFB

VVTFB

1None

Jointangle

error

Average

subjecterror

(RMSE)

VVT=overall#errorof

21%

(sig)

Hingejointerrors

#(sig),rotationaljoint6¼

# 15%"perform

ance,

7%

accelerate

learning

Sigristetal26

Randomisedcontrolled

trial

F:8;M:28

Total:36

28±3.7

(22–40)

Rowing

Lab

Concurrent

VFB

Concurrent

AFB

Concurrent

HFB

Term

inalFB

(control)

3Short

term

:start

days2and

3 Longterm

:1day

1week

Temporal

error

Spatialerror

RMSE

Error

intiming

Errorin

velocityratio

Movement

variability

H=larger+",

movements

(sig)FB+

nFBV="perform

ance

withFB+#cognitive

demanding

Tearly"perform

ance,

but"temporalerror

H=perform

anceFB>

nFB(fasterexecution

innFB)

H=#temporaland

velocityerror

Term

inalFB="

effectivethan

concurrenttypeofFB

Spelm

ezanetal29

Trial

1Exploratory

study

F:8;M:12

Total:20

25.25

(22–28)

—Lab

Tactile

stimulation†

1None

Description

ofperceived

FB

""Variety

inresponses

(50/50:pushvspull)

SP=slower

processing+"

attention

Trial

2Clinicaltrial

F:4;M:14

Total:18

(19–30)

Wii-balance

board

(slalom)

Lab

Random

tactile

motion

instructions

11day

%Pattern

recognition

High%

pattern

recognitionrelaxed+

mobile

Nosig

,perform

ance

retentionversustest

Trial

3Clinicaltrial

F:4;M:6

Total:10

(23–28)

Snowboarding

Reality

based

VTFB*

Audio

instructions

1None

RT

%Pattern

recognition

Perception

audio

instructions

FasterRTto

tactile

instructions

Audio

instructions

easierto

interpret

%RecognitionA>VT

(97%

vs87%

)

Con

tinued

6 van Breda E, et al. BMJ Open Sport Exerc Med 2017;3:e000216. doi:10.1136/bmjsem-2016-000216

Open Access

Table

4Continued

Study

Design

Study

population

(n)

Mean

age

(years)

Movement

orbranchof

sports

Setting

Feedback

conditions

Training

(Noof

days)

Retention

Outcome

measures

Main

results

28

Randomisedclinicaltrial

F:22

20.1±1.06

Rowing

Lab

DirectVT

positionalFB

Directn-

positionalFB

Delayedn-

positionalFB

1None

HR

Tim

ingof

knee/back

angle

Speed

VT:#HRovertests

("in

otherFBgroups)

Nosig

,in

speedand

timingbetw

eenFB

groups

Spelm

ezan23

Clinicaltrial

F:2;M:8

Total:10

(21–29)

Snowboarding

Reality

based

Verbal

instructions*

Verbal+VTFB

1None

System

recognition

accuracy

M(O

)%

Riding

mistakes

Verbal="effectivefor

learningnew

skills

VTFB+verbalFB="

cognitiveload

VTFB=effectivein

sim

ple

exercises

Jansenetal24

Clinicaltrial

F:7;M:8

Total:15

(19–25)

Wristrotation

Lab

SRT*

CRT

1None

RT

RTforL+Rrotations

EFR=mostintuitive(=

push)+#RT

vanderLinden

etal27

Randomisedcontrolled

trial

F:3;M:5

Total:8

25–30

Playingthe

violin

Reality

based

FB

Control

4Minim

um

1day

Proxim

al-

distal

deviation

Vertical

deviation

Sig

,pretestversus

test(VT:#proxim

al-

distaldeviation)

n-sig:testversuspost-

testfi

nolearning

effectC

andFB

Control:noobvious

overallprogress

25

Clinicaltrial

F:15;M:11

Total:26

23.2

(18–

49)

1–3dfarm

motions

Lab

VFB*

VVTFB

44days

RMSangle

error

Arm

angle

error

Motionerror

AverageRMSEVVT=

#for1dfmotion(sig)

VVT:larger#errorday

4(1

and3df)

VTFBdifficultto

interpretin

multiple

df

WorkloadVVT>V(but

sig

#overtraining)

Nosig

effectofFBin

retentionorprobetrial

*Within-subjectdesign.

†Betw

een-subjects

design.

AFB,auditory

feedback;CRT,choicereactiontime;df,degreesoffreedom;EFR,extrinsic

frameofreference;F,female;FB,feedback;HFB,hapticfeedback;HR,heartrate;L,left;M,male;

M(O

),systemslatency;nFB,non-feedbacktrials;n-positional,non-positional;n-sig,non-significant;R,right;RMS,rootmeansquare;RMSE,rootmeansquare

error;RT,reactiontime;sig,

significant;SP,sim

ultaneouspatterns;SRT,sim

ple

reactiontime;VFB,visualfeedback;VT,vibrotactile;VTFB,vibrotactile

feedback;VVT,visualandvibrotactile;VVTFB,visualandvibrotactile

feedback.

van Breda E, et al. BMJ Open Sport Exerc Med 2017;3:e000216. doi:10.1136/bmjsem-2016-000216 7

Open Access

to analyse the data from the included studies in afashion as has been proposed by Saw et al.20 However,because of the diversity in outcome parameters and inobjective data, we have not been able to perform suchanalysis. Therefore, the results of the individual studieswill be presented qualitatively.

RESULTSStudy selectionThe PRISMA flow diagram (figure 1) summarises ourstudy selection process. The initial search yielded 312results, of which 16 duplicates were deleted, leaving a

total of 296 studies. The manual search and repeatedsearch yielded a total of 25 extra studies, making atotal of 321 studies that were selected for screening.From the initial 296 studies, 274 were excluded basedon title and abstract. Subsequently, 38 studies wereexcluded based on full-text screening. Of the 25studies obtained by the manual and repeated searchperformed on 30 March 2017, a total of 14 wereexcluded based on title and abstract, and 6 on the full-text screening. Nine9 studies remained eligible for thisreview.

Table 5 Methodological scoring using the PEDro checklist and level of evidence of individual studies according to the

CBO guidelines

Study

Items Methodological

quality

Level of

evidence1* 2 3 4 5 6 7 8 9 10 11 Score**

Eid et al21 � ? ? ? � � � � + � � 1/10 Poor B

Lieberman and Breazeal22 � ? ? ? + ? ? + + + + 5/10 Fair B

Sigrist et al26 + + ? + ? ? ? + + + + 6/10 High B

Spelmezan et al29 Study

1

� + ? + + ? ? ? + � � 3/10 Poor C

Study

2

� � � � ? ? ? + + + + 4/10 Fair C

Study

3

� + ? � ? � ? + + + + 5/10 Fair B

Van Erp et al28 ? + ? + ? ? ? + + + + 6/10 High B/A2

Spelmezan23 � + ? � � + + + + + + 7/10 High B

Jansen et al24 ? � � + ? ? ? + + + + 5/10 Fair B

van der Linden et al27 � � ? + ? ? ? + + + + 5/10 Fair B

25 � + ? + ? ? ? + + + + 6/10 High B

Scoring by the PEDro scale:

+ = Yes

� = No

? = Too little information or doubtable.

Legend:

1: Eligibility criteria were specified.

2: Subjects were randomly allocated to groups (in a crossover study, subjects were randomly allocated an order in which treatments were

received).

3: Allocation was concealed.

4: The groups were similar at baseline regarding the most important prognostic indicators.

5: There was blinding of all subjects.

6: There was blinding of all therapists who administered the therapy.

7: There was blinding of all assessors who measured at least one key outcome.

8: Measures of at least one key outcome were obtained from more than 85% of the subjects initially allocated to groups.

9: All subjects for whom outcome measures were available received the treatment or control condition as allocated or, where this was not

the case, data for at least one key outcome were analysed by ‘intention to treat.’

10: The results of between-group statistical comparisons are reported for at least one key outcome.

11: The study provides both point measures and measures of variability for at least one key outcome.

PEDro scores:

High quality = score 6–10

Fair quality = score 4–5

Poor quality = score �3

*Item 1 pertains for the external validity, thus is not considered for the final score.

**Poor quality = PEDro score �3.

CBO, Dutch Institute for Healthcare Improvement; PEDro, Physiotherapy Evidence Database.

8 van Breda E, et al. BMJ Open Sport Exerc Med 2017;3:e000216. doi:10.1136/bmjsem-2016-000216

Open Access

The majority of the studies were excluded based onintervention and population, 63% and 20%, respec-tively. Twenty20 studies were excluded based on itsstudy design and 24 based on outcome. Finally, three3

studies were excluded based on subject, eight8 studieson language and four4 studies were excluded based onage of the population. The majority of full-text exclu-sions were based on intervention.

Study characteristicsFor each included study, the characteristics of dataextraction are shown in table 4. All nine9 studies werepublished in English and could be designated as clin-ical study designs with VTFB as (one of) theintervention(s). The majority of the studies involvedsimple clinical trials21–25 whereas two studies26 27

involved a randomised controlled trial. Only one studycould be designated as a randomised clinical trial.28

The study of Spelmezan et al29 consists of one explor-

atory clinical trial with an open response paradigm andtwo clinical trials, one of which was performed in areality-based context. Two studies investigated theeffects of VTFB during rowing, two during snow-boarding, one focused on treadmill running and twoduring a simple arm movement.26 28 One study investi-gated the effects of VTFB during an artistic skill.27

In two studies23 27 and in one of the trials of Spel-mezan et al,29 the experiments were done in a realrather than a laboratory setting.VTFB was compared with other FB modalities in the

majority of studies, except for the study of Jansenet al

24 and in the first two trials of Spelmezan et al.29 Inthe latter, VTFB was the only intervention. Outcomeparameters differed between all studies. The mainoutcome parameters were joint angle error, root meansquare error (RMSE), percentage of correct patternrecognition, reaction time and HR. Secondary andadditional outcome parameters included temporal andspatial errors, timing errors and subjective descriptionof perceived FB. In five studies,22 23 25 26 29 partici-pants’ subjective impression on the FB modalities wasassessed using a questionnaire or interview. Durationof the intervention, that is, number of training days,ranged from 1 to 4days. Only four studies included aretention phase ranging from a 1-day retention test toa 1-week26 retention test.

Risk of bias within studiesTable 5 shows the methodological scoring and theassessment of the level of evidence.We found a significant Cohen’s kappa agreement

between the two authors who performed the reviewingconcerning the scoring of methodological quality(kw=0.983; p<0.001). In 97% of the cases, bothreviewers agreed. In all cases, disagreements wereresolved by consensus, and if no consensus could bereached by the two authors (KW and TP) a third authorwas asked (SV). Each item contributes one point to the

total PEDro score ranging from 0 to 10 points. Item 1is not considered for the final score as it relates toexternal validity. Of the nine included studies, one hada final PEDro score of 1,21 three22 24 27 had a finalscore of 5, three25 26 28 had a final score of 6 andone23 had a final score of 7. The study of Spelmezanet al

29 contains three trials of which one had a finalscore of 3, one had a final score of 4 and the third hada final score of 5. The overall quality of the includedstudies was fair to excellent, with the exception of thestudy of Eid et al (score 1)21 and the first trial of theSpelmezan et al’s study (score 3).29 Most studies wererated with a lower score based on blinding andrandomisation.

Results of individual studiesThe effects of VTFB in regulating optimal training intensityEid et al studied the effectiveness of VT stimuli toadjust exercise intensity.21 A within-subject design wasused to compare ‘training with VTFB’ with no deviceand one with a Nike+ SportBand, using HR as themain outcome parameter. Unfortunately, only onesubject was presented, but based on VTFB was shownto be more efficient to regulate HR, showing fewerfluctuations and resulting in a smooth change inmethod by taking bigger steps and breathing moreefficiently. A faster HR recovery was found in theVTFB condition.

VTFB in learning single and multiple degrees of freedom ofarm movements and initiating wrist rotationsLieberman and Breazeal22 and Bark et al

25 studiedwhether VTFB, based on the push protocol andsensory saltation, accelerates learning of single andmultiple degrees of freedom (df) arm movements andreduces motion errors. Lieberman22 compared visualFB (n=20) with the combination of visual FB andVTFB (n=20). They found that the addition of VTFBto motor training induces significant change in perfor-mance: a reduction in real-time errors by up to 27%and an improvement of learning rate by up to 23%;with a steady state in learning errors by 27% overflexion joints. Subjects with the combinationFB showed a significant higher level of correcting theirmotions in comparison to those without TCB. Nosignificant loss of comfort through the addition of thewearable has been reported by the subjects.In line with the studies of Lieberman et al,22 Bark

et al25 used a within-subject design to compare visual

FB with the combination of visual FB and VTFB. Theyreported that RMSE was significantly lower for thecombination group (15.7�

±6.5�) in comparison to thevisual only group (17.1�

±6.4�) for learning 1 df motion,but not for 2 or 3 df motions. In both groups, motionerrors decreased over the 4 days of training but nosignificant effect on joint angle error was found duringprobe or retention.

van Breda E, et al. BMJ Open Sport Exerc Med 2017;3:e000216. doi:10.1136/bmjsem-2016-000216 9

Open Access

Finally, Jansen et al studied whether VTFB was usefulin initiating wrist rotations and which stimulus locationprovided the fastest reaction time.24 They concludedthat an extrinsic reference frame in which the vibratorysignal ‘pulls’ the hand in the desired direction is mostintuitive.

Effectiveness of VTFB in a real sports contextSpelmezan et al conducted a within-subject study insnowboarding in which traditional verbal instructionsin combination with VTFB were compared with tradi-tional verbal FB only.23 Verbal instructions in themorning resulted in fewer mistakes compared withVTFB (p=0.044), whereas the opposite was true fortrials in the afternoon (p=0.014). Verbal instructionsoutweighed VTFB when practising new tasks. Suchresults are in line with another study of Spelmezan et al

in which snowboarders were provided with verbalFB instructions with or without VTFB.29 A fasterresponse time has been found for VTFB in comparisonwith verbal FB (p=0.01).The effectiveness of VTFB in learning a rowing task

was studied by Van Erp et al and Sigrist et al.26 28 VanErp et al

28 compared direct positional VTFB withdirect non-positional verbal FB and delayed non-posi-tional verbal feedback. Neither group showed alearning effect. The VT group showed a significant(p<0.02) decrease in HR over tests but no differencewas found between groups regarding timing andspeed. Sigrist et al

26 compared VTFB with verbalFB and auditory FB with a self-controlled terminal FB.Terminal self-controlled FB was found to be superiorto all types of concurrent AFB. Furthermore, a signifi-cant decrease in RMSE, temporal error and error invelocity ratio was found from baseline to retention indi-cating that VTFB is effective in enhancing temporalaspects, rather than spatial aspects of movement.Again, a guidance effect could be found, that is,decrease in performance in non-FB conditions.30

DISCUSSIONThe aim of the present review was to provide a moredetailed insight in the scientific literature concerningVTFB as a tool for sports performance enhancement.Only nine studies, based on the methodological criteriaset, were eligible for this review. The quality of thestudies was fair but the level of evidence was subop-timal. We conclude that, although VTFB has beenshown to be effective under some conditions, evidencefor its use as sports performance enhancement remainsscarce and challenging.10 11 31

Summary of evidenceVTFB has been used in navigation,32 aviation33 and inautomotive.34 It also found its way in virtual reality,gaming and in art.35 36 In healthcare, VTFB has beenused as a navigation tool in the visually

impaired subjects, as sensory substitution in the reha-bilitation of patients with stroke, and in vestibularimpaired subjects where it has been used as sensoryaugmentation.37 To the best of our knowledge, its usein sports practice is very limited. A reason for this isnot easy to provide but the limited number of scientificstudies in concordance with the broad range ofoutcome parameters and variability of the results inlaboratory setting may be the major cause of the lack ofuse in sports practice.Out of the nine studies eligible for final interpreta-

tion, only the studies of Spelmezan et al and van derLinden et al

23 27 29 addressed VTFB in a reality-basedcontext. For instance, van der Linden et al imple-mented the use of VTFB in real training situations forartists (violin players) whereas Spelmezan used snow-boarders as study population. Subjects in the laterstudy, however, stated that the VTFB instructionshindered normal movement because VT instructionswere randomly provided in time and the placement onthe body/limbs was reported to be frustrating. Thismight be one of the major reasons for athletes not touse VTFB devices on their body/limbs even if it mighthelp them to improve performance. These findings arein contrast with the findings of Lieberman et al,22 whoreported no loss of comfort during the application ofVTFB. The use of devices that differ in size, weightand position could play a major role in the acceptanceof VTFB in sports practice. Future studies shouldtherefore be designed such that VTFB does not inter-fere with normal movement by providing VTinstructions on suitable moments of VTFB.There is some evidence that VTFB is effective in

regulating exercise intensity during physical perfor-mance by keeping HR within aerobic limits.18 Byapplying VTFB, athletes could, based on their maximalHR, be guided by tactile information to perform at aless intense level and thus exercise more efficiently.Exercising with VTFB resulted in both short term andlong term, in a decrease in HR.22 The athletes in theirstudy who received TFB (ie, direct and positional)showed a decreased HR in the post-test as comparedwith the pretest whereas the athletes who did not getTFB showed an increase in HR. A plausible explana-tion for this HR effect was that athletes without VTFBwere less efficient in maintaining motor skills andmore than necessary adjustments in motor controlwere necessary. Therefore, it is reasonable to assumethat VTFB is effective in improving temporal move-ment aspects, but not spatial errors.26 These findingssupport the results of previous studies.38 39 Lastly,there is moderate evidence that VTFB instructionscould support the athlete in practising and fine-tuninga previously learnt movement. A second explanationfor an over-the-day decrease in HR is that VTFB is lessmental demanding.25 To date, no studies are presentthat support this idea.

10 van Breda E, et al. BMJ Open Sport Exerc Med 2017;3:e000216. doi:10.1136/bmjsem-2016-000216

Open Access

Some studies used joint angle errors as the mainoutcome parameter.19 22 This is somewhat surprisingsince joint angle is not necessarily the most importantoutcome parameter for motor performance, and assuch, athletic performance. Moreover, during a motortask, several joint angles should be taken into account,that when applying VTFB should provide simultaneousand multiple VTFB signals.23 25 27 The major draw-back, however, of some studies using VTFB as errorsignal is the lack of implementing an order orsequence in signals.22 28 Subjects are continuouslyprovided with FB signals which lead to overcorrectionof motor tasks/movements. Only a few scientific studiesthat used VTFB for motor task correction used moreaccurate and timed signals.21 25–27

Limitations of the current reviewThe quality and diversity of the studies made it impos-sible for us to do quantitative analysis of the results.The risk of bias of individual studies was overallmoderate to fair and because randomisation was inade-quate or lacking in the majority of trials, level ofevidence was moderate to weak. Furthermore, poorand incomplete reporting of study designs, patientpopulations and interventions made the interpretationand synthesis of the included studies much more diffi-cult and may have contributed to publication bias.Therefore, generalisation of the results to athletes isdifficult.Finally, the main limitation of this study is that the

context of VTFB and comparison between different FBmodalities and outcome measures for performancewere not the same across the studies, making a directcomparison between the outcome of these studiesdifficult.

CONCLUSIONSOur review suggests that VTFB, which has been used asapplication in medical and non-medical-based contexts(ie, patient groups, automotive, and so on), lacks scien-tific support for enhancement of sports performanceoutcomes. The inconsistency of findings and moderatelevel of support, as reported in the present study,hardly provide evidence to support the suspected role.More basic and applied studies are necessary towarrant its use in a sports context.

Contributors The study was planned by EvB, SV and ST. KW and TPconducted the literature search whereas SV was used as author in case ofdisagreements during selection. EvB and SV wrote the initial draft of themanuscript. EvB, SV, WS and ST discussed and revised the manuscripttogether and gave final approval for publication.

Competing interests None declared.

Provenance and peer review Not commissioned; externally peer reviewed.

Open Access This is an Open Access article distributed in accordance withthe Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license,which permits others to distribute, remix, adapt, build upon this work non-

commercially, and license their derivative works on different terms, providedthe original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/

Correction notice This paper has been amended since it was publishedOnline First. Owing to a scripting error, some of the publisher names in thereferences were replaced with 'BMJ Publishing Group'. This only affected thefull text version, not the PDF. We have since corrected these errors and thecorrect publishers have been inserted into the references.

© Article author(s) (or their employer(s) unless otherwise stated in the text ofthe article) 2017. All rights reserved. No commercial use is permitted unlessotherwise expressly granted.

REFERENCES

1. Schmidt RA, Wrisberg CA. Motor learning and performance: asituation-based learning approach: human kinetics, 2008.

2. Utley A, Astill S. Motor control, learning and development. Int JSports Sci Coach 2008;3:297–9.

3. Sch€onauer C, Fukushi K, Olwal A et al, eds. Multimodal motionguidance: techniques for adaptive and dynamic feedback.Proceedings of the 14th ACM International Conference onMultimodal Interaction: ACM, 2012.

4. Minogue J, Jones MG. Haptics in Education: Exploring an UntappedSensory Modality. Rev Educ Res 2006;76:317–48.

5. Sigrist R, Rauter G, Riener R, et al. Augmented visual, auditory,haptic, and multimodal feedback in motor learning: a review.Psychon Bull Rev 2013;20:21–53.

6. El Saddik A, Orozco M, Eid M, et al. Haptics: generalprinciples Haptics technologies. Springer, 2011.

7. Hale KS, Stanney KM. Deriving haptic design guidelines from humanphysiological, psychophysical, and neurological foundations. IEEEComput Graph Appl 2004;24:33–9.

8. Lauber B, Keller M. Improving motor performance: selected aspectsof augmented feedback in exercise and health. Eur J Sport Sci2014;14:36–43.

9. Sch€atzle S, Ende T, W€usthoff T et al, eds . Vibrotac: an ergonomicand versatile usable vibrotactile feedback device. RO-MAN: 2010IEEE, 2010.

10. Alahakone AU, Senanayake S, Arosha M, eds. Vibrotactile feedbacksystems: current trends in rehabilitation, sports and informationdisplay. Advanced Intelligent Mechatronics, 2009 AIM 2009 IEEE/ASME International Conference on 2009: IEEE, 2009.

11. Choi S, Kuchenbecker KJ. Vibrotactile display: Perception,technology, and applications. Proceedingsof the IEEE.2013;101:2093–104.

12. Gray R. Multisensory Information in the Control of Complex MotorActions. Curr Dir Psychol Sci 2008;17:244–8.

13. Amemiya T, Ando H, Maeda T. Lead-me interface for a pullingsensation from hand-held devices. ACM Trans Appl Percept2008;5:1–17.

14. Spelmezan D, Ziefle M. A language of tactile motion instructions forphysical activities. RWTH Aachen University, 2011.

15. Moher DLA, Tetzlaff J, Altman DG. Preferred Reporting Items forSystematic Reviews and Me.Analyses: The PRISMA Statement,2009. http://prisma-statementorg/.

16. Pressley M, Ghatala ES. Metacognitive benefits of taking a test forchildren and young adolescents. J Exp Child Psychol1989;47:430–50.

17. Rodrigue KM, Kennedy KM, Raz N. Aging and longitudinal change inperceptual-motor skill acquisition in healthy adults. J Gerontol BPsychol Sci Soc Sci 2005;60:P174–P181.

18. de Morton NA. The PEDro scale is a valid measure of themethodological quality of clinical trials: a demographic study. Aust JPhysiother 2009;55:129–33.

19. Keus SH, Bloem BR, Hendriks EJ, et al. Practice RecommendationsDevelopment G. Evidence-based analysis of physical therapy inParkinson’s disease with recommendations for practice andresearch. Mov Disord 2007;22:451–60.

20. Saw AE, Main LC, Gastin PB. Monitoring the athlete trainingresponse: subjective self-reported measures trump commonly usedobjective measures: a systematic review. Br J Sports Med2016;50:281–91.

21. Eid M, Saad U, Afzal U, eds. A real time vibrotactile biofeedbacksystem for optimizing athlete training. Audio Visual environments andGames (Have), 2013 IEEE International Symposium on 2013. IEEE.

22. Lieberman J, Breazeal C. TIKL: development of a wearablevibrotactile feedback suit for improved human motor learning. IEEETransactions on Robotics 2007;23:919–26.

van Breda E, et al. BMJ Open Sport Exerc Med 2017;3:e000216. doi:10.1136/bmjsem-2016-000216 11

Open Access

23. Spelmezan D, ed An investigation into the use of tactile instructionsin snowboarding. Proceedings of the 14th international conference

on Human-computer interaction with mobile devices and services.

ACM;2012.24. Jansen C, Oving A, van Veen H-J. eds. Vibrotactile Movement

Initiation.Proceedings of Eurohaptics. Citeseer, 2004.25. Bark K, Hyman E, Tan F, et al. Effects of Vibrotactile feedback on

human learning of arm motions. IEEE Trans Neural Syst Rehabil Eng

2015;23:51–63.26. Sigrist R, Rauter G, Riener R, et al. Terminal feedback outperforms

concurrent visual, auditory, and haptic feedback in learning a

complex rowing-type task. J Mot Behav

2013;45:455–72.27. van der Linden J, Schoonderwaldt E, Bird J, et al. Musicjacket—

combining motion capture and vibrotactile feedback to teach violin

bowing. IEEE Trans Instrum Meas 2011;60:104–13.28. Van Erp JB, Saturday I, Jansen C, eds. Application of tactile displays

in sports: where to, how and when to move. Proc Eurohaptics.

Citeseer;2006.29. Spelmezan D, Jacobs M, Hilgers A, eds. Tactile motion instructions

for physical activities. Proceedings of the SIGCHI conference on

human factors in computing systems. ACM, 2009.30. Salmoni AW, Schmidt RA, Walter CB. Knowledge of results and

motor learning: a review and critical reappraisal. Psychol Bull

1984;95:355–86.

31. Shull PB, Damian DD. Haptic wearables as sensory replacement,sensory augmentation and trainer–a review. J Neuroeng Rehabil2015:12 1.

32. Erp JBFV, Veen HAHCV, Jansen C, et al. Waypoint navigation with avibrotactile waist belt. ACM Trans Appl Percept 2005;2:106–17.

33. Rupert AH. Tactile situation awareness system: proprioceptiveprostheses for sensory deficiencies. Aviat Space Environ Med2000;71(Suppl 9):A92–9.

34. Scott JJ, Gray R. A comparison of tactile, visual, and auditorywarnings for rear-end collision prevention in simulated driving. HumFactors 2008;50:264–75.

35. Feeling the beat where it counts: fostering multi-limb rhythm skillswith the haptic drum kit. In: Holland S, Bouwer AJ, Dalgelish M, eds.Proceedings of the fourth international conference on Tangible,embedded, and embodied interaction: ACM, 2010.

36. Mobile music touch: mobile tactile stimulation for passive learning.In: Huang K, Starner T, Do E, eds. Proc SIGCHI Conf Hum FactorComput Syst : ACM, 2010.

37. Wall C, Weinberg MS, Schmidt PB, et al. Balance prosthesis basedon micromechanical sensors using vibrotactile feedback of tilt. IEEETrans Biomed Eng 2001;48:1153–61.

38. Crespo LM, Reinkensmeyer DJ. Haptic guidance can enhance MotorLearning of a steering Task. J Mot Behav 2008;40:545–57.

39. Haptic interfaces for virtual environment and teleoperator systems,2002 haptics 2002 proceedings 10th symposium on; 2002: IEEE. In:Feygin D, Keehner M, Tendick F, eds. Haptic guidance: Experimentalevaluation of a haptic training method for a perceptual motor skill.

12 van Breda E, et al. BMJ Open Sport Exerc Med 2017;3:e000216. doi:10.1136/bmjsem-2016-000216

Open Access