Clinimetric Properties of the Timed

Up and Go Test for Patients With

Stroke: A Systematic Review
Thra B. Hafsteinsdttir,1,2,3 Marijke Rensink,3 and Marieke Schuurmans1,2,3
1

Department of Rehabilitation, Nursing Science and Sport, Rudolf Magnus Institute, University Medical Center Utrecht, Utrecht,
The Netherlands; 2Clinical Health Sciences, Faculty of Medicine, Utrecht University, The Netherlands; 3The Research Centre for
Innovations in Health-Care, University of Applied Sciences Utrecht, Utrecht, The Netherlands

Objective: To systematically review and summarize the clinimetric properties, including reliability, validity, and
responsiveness, the procedures used, and the meanings of the scores in the Timed Up and Go Test (TUG). The TUG is a
performance test that identifies problems with functional mobility in patients with stroke. Methods: MEDLINE and the
Cochrane Central Register of Controlled Trials were searched from 1991 to January 2013. Studies were included if (1) the
participants were adults with stroke; (2) the research design was cross-sectional, descriptive, or longitudinal and examined
the clinimetric properties, including reliability, validity, and sensitivity to change, and procedural differences in the TUG;
and (3) the study was published in English from 1991 to January 2013. Results: Thirteen studies met the inclusion criteria.
Of these, 4 showed the TUG to have good convergent validity, as it had significant correlations with various instruments.
Three studies that investigated the test-retest reliability showed the TUG to have excellent intrarater and interrater reliability
(intraclass correlation coefficient [ICC] > 0.95). The 3 studies that investigated whether the TUG could predict falls after
stroke showed inconclusive results. Three studies showed the TUG to be sensitive to change, and 1 study showed the TUG
to be responsive in moderate- and fast-walking patients with stroke. However, there were wide variations in the procedures
and instructions used. Conclusion: The TUG can be recommended for measuring basic mobility skills after stroke in patients
who are able to walk. However, the procedures and instructions should be described more clearly. Key words: functional
mobility, psychometric properties, rehabilitation, reliability, stroke, Timed Up and Go test, validity

ost patients with stroke are confronted

with motor disabilities and balance
problems. Functional mobility at home
and in the community requires a certain gait speed
and the ability to maintain balance during walking
and turning.1 Patients with stroke, who also have
problems with balance, have a high risk for falling, and the prevalence of falls can be as high as
73%.2 Therefore, an important rehabilitation goal
after stroke is maintaining the ability to walk at
home, in and outside of the house, and in the community. Health care professionals in care facilities
provide interdisciplinary rehabilitation training for
patients to improve basic functional mobility. This
training generally involves several components:
assessment to identify and quantify the patients
needs, goal setting to define goals for improveCorresponding author: Thra B. Hafsteinsdttir, PhD, Associate
Professor, University Medical Center Utrecht, The Department of
Rehabilitation, Nursing Science and Sport, PO Box 85500, 3508 GA
Utrecht, The Netherlands; phone: 00-31-(0)6 1007 5163; e-mail:
t.hafsteinsdottir@umcutrecht.nl

ment, intervention to assist in achieving goals, and

reassessment to evaluate progress toward agreedupon goals.3 The Timed Up and Go Test (TUG) is
a widely used performance test for the evaluation
of functional mobility or basic mobility skills in
elderly people.4,5 The TUG is easy to administer
compared with other performance measures and
provides information on the abilities that facilitate
living safely at home. The TUG requires participants to stand up from a chair, walk 3 meters, turn
around, return to the chair, and sit down again.
The time required to complete the test is recorded
in seconds using a stopwatch. A participant may
walk with a cane or other walking aid.4
When the usefulness of the TUG in the
rehabilitation of patients with stroke is considered,
it is important to know the clinimetric properties,

Top Stroke Rehabil 2014;21(3):197210

2014 Thomas Land Publishers, Inc.
www.strokejournal.com
doi: 10.1310/tsr2103-197

197

198

TOPICS IN STROKE REHABILITATION/MAY-JUNE 2014

including validity and reliability, as well as

the interrater and intrarater reliability. Studies
investigating the clinimetric properties of the
TUG for patients in geriatric units demonstrated
excellent interrater and intrarater reliability,
a satisfactory construct validity, and good
responsiveness to change.6-8 Although the TUG is
widely used to evaluate the basic mobility skills of
patients with stroke, no overview of the clinimetric
properties of the TUG within the stroke population
is available.
The aim of this review was to compile an
overview of the literature on the clinimetric
propertiesincluding the reliability, validity,
responsiveness, and sensitivity to change of the
TUGin patients with stroke. In addition, we also
reviewed the ability of the TUG to predict falling,
the standardization of the test, the normative
scores, and the procedures used.
Methods
The literature was systematically reviewed
according to the steps of the Preferred Reporting
Items for Systematic Review and Meta-Analysis
(PRISMA) statement.9
Search strategy

The following databases were searched for

entries from January 1, 1991, through January
31, 2013: MEDLINE (PubMed) and the Cochrane
Central Register of Controlled Trials. The
databases were searched for relevant studies using
the following key words: Timed Up and Go,
Timed Up and Go Test, TUG, and stroke (MESH
term) in combination with the following terms:
psychometric properties, clinimetric properties,
validity, reliability, responsiveness, falls, and
falling. Reference lists of relevant articles were
searched to supplement the studies identified in
the electronic search.
Selection criteria

Studies were selected for inclusion if (1) all

participants were adults (18 years of age) who
had sustained a stroke; (2) the research design was
cross-sectional, longitudinal, or descriptive and

examined the clinimetric properties, including the

reliability, validity, and sensitivity to change of the
TUG, and the reliability and validity of balance
and mobility tests compared with the TUG; (3) the
study was published after 1991, when the original
article by Posiadlo and Richardson4 was published,
and before January 2013; and (4) the publication
language was English.
Study selection and methodological
quality assessment

The titles and abstracts of all the articles identified

in the literature were screened for eligibility. The
titles were initially screened by 1 reviewer (M.R.),
and subsequently, 2 independent reviewers (M.R.
and T.B.H.) screened the relevant abstracts and
reviewed the full text of selected articles. The
following characteristics were extracted into a
data extraction form: author, publication year,
study setting, sample, aim of the study, design,
clinimetric properties measured, description of the
used variation of the test, and results.
Although a critical quality appraisal is a
fundamental component of systematic reviews,
validated appraisal criteria evaluating the
clinimetric properties of instruments are scarce.
The studies were independently evaluated by 2
reviewers (M.R. and T.B.H.) with the following
validated quality criteria, which had excellent
preconsensus interrater reliability (intraclass
correlation coefficient [ICC], 0.82-0.91) 10,11:
(1) thorough literature review to define the research
question; (2) specific inclusion/exclusion criteria;
(3) specific aim/purpose and question; (4) at least
1 psychometric property reported; (5) sample
size calculation/justification; (6) appropriate
retention/follow-up; (7) references by authors to
specific administrative procedures, scoring, and
interpretation of procedures; (8) standardized
measurement techniques; (9) data presented
for each question; (10) appropriate statistics
point estimate; (11) appropriate statistical error
estimates; and (12) valid conclusions and clinical
recommendations. If an item was fulfilled, it was
rated as 1; if an item was not fulfilled, it was
rated as 0; and if an item was not applicable,
it was rated as N/A. A score of 9 to 12 was
considered good quality, a score of 6 to 8 was

Timed Up and Go: Clinimetric Properties After Stroke

moderate quality, and a score of 5 was poor

quality. The quality assessment is presented in
Table 1.

used; and (4) TUG scores of the patients with

stroke. Descriptions of the characteristics of the
studies, including the clinimetric properties, are
presented in Table 2, and the procedures and
scores of the TUG are presented in Table 3.

Results
The initial search identified 179 studies relevant
to the aims of the review. The titles and abstracts
of these studies were screened if they met the
inclusion criteria, and 33 studies were selected.
After the full text of these articles had been
read, 13 studies met the inclusion criteria and
were included in the review.5,12-23 A flow chart
of the study selection procedure is presented in
Figure 1. Because of the wide variety of study
designs and methods used, it was not possible
to conduct a meta-analysis, and therefore the
results are presented in a narrative manner in
the following sections: (1) characteristics of
the studies; (2) clinimetric properties divided
according to validity, reliability, responsiveness,
and predictive value; (3) application procedures

Table 1.

199

Characteristics of the studies

Five of the included studies in which the

clinimetric properties of the TUG are compared
with other measurements are descriptive cohort
studies.5,12-14,23 The 4 studies that investigated the
usefulness of the TUG for patients with stroke
had a cross-sectional design.15-18 Three studies
investigated the ability of the TUG to predict falls.
Of these, 1 study had a cross-sectional design,19
and 2 had a prospective descriptive design.20,21
The studies involved a total of 673 patients with
stroke. Because the TUG requires the participants
to walk independently, in 12 of the studies, the
patients were able to walk 10 meters unassisted,
with or without walking aids. In one study,

Methodological quality of the studies included

Author, year

10

11

12

Faria et al., 2012

Persson et al., 2011
Simpson et al., 2011
Knorr et al., 2010
Hollands et al., 2010
Faria Coelho de Morais et al., 2009
Heung and Ng, 2009
Jonsdottir et al., 2007
Belgen et al., 2006
Andersson et al., 2006
Flansbjer et al., 2005
Ng and Hui-Chan, 2005
Salbach et al., 2001

1
1
1
1
1
1
1
1
1
1
1
1
1

1
1
1
1
1
1
1
1
1
1
1
1
1

1
1
N/A
1
N/A
1
1
1
1
1
1
1
1

1
1
1
1
N/A
N/A
N/A
1
1
1
1
1
N/A

1
0
0
0
1
1
1
1
0
0
0
1
0

1
1
1
1
N/A
N/A
N/A
N/A
N/A
1
N/A
N/A
1

1
1
1
1
1
1
1
1
1
1
1
1
1

1
1
1
1
1
1
1
1
1
1
1
1
1

1
1
1
1
1
1
1
1
1
1
1
1
1

1
1
1
1
1
1
1
1
1
1
1
1
1

1
0
0
0
0
1
0
1
0
0
1
0
1

1
1
1
1
1
1
1
1
1
1
1
1
1

Note: 1= item fulfilled; 0 = item not fulfilled; N/A= item not applicable to study.
1. Thorough literature review to define research question
2. Specific inclusion/exclusion criteria
3. Specific hypotheses (or specific objective/question/aim/purpose)
4. Appropriate scope of psychometric properties (at least 1 psychometric property reported)
5. Sample size calculation/justification
6. Appropriate retention/follow-up
7. Authors referenced specific procedures to administer, score and interpret the procedures
8. Measurement techniques were standardized (within the study)
9. Data were presented for each hypothesis (or question)
10. Appropriate statistics-point estimate
11. Appropriate statistical error estimates
12. Valid conclusions and clinical recommendations

200

TOPICS IN STROKE REHABILITATION/MAY-JUNE 2014

179 papers identified in data base searches

PUBMED
175 papers

Cochrane
4 papers

After screening titles

167 papers

After screening abstracts

30 papers
1 paper found by hand
18 papers excluded
13 studies included

Figure 1. Flow chart of the selection of studies.

patients who were both able and unable to walk

were included.21 In 4 studies, the TUG scores
of patients with stroke were compared with the
scores of healthy, elderly people.15,17,18,22
Regarding the time and setting, one study
started in the acute phase with reassessments
after 3, 6, and 12 months.19 Three studies were
conducted in the subacute phase and recruited
patients from a geriatric day care unit,16 a center for
subacute care,12 and a day hospital.5 Nine studies
were conducted in the chronic phase, with start
times varying from 4 weeks after stroke22 to more
than 1 year after stroke.15,23
In terms of location, 3 studies were conducted
in Canada,12,14,22 3 in Sweden,13,20,21 2 in Hong
Kong,15,16 2 in Brazil,17,23 one in the United
Kingdom,18 1 in Italy,5 and 1 in the United States.19
All 13 studies were judged to be of good quality
(scoring 10 or more out of a possible 12 points).
Clinimetric properties

Validity

Construct validity was investigated as convergent

validity in 4 studies.5,13-15 The association between
the scores on the TUG and ankle spasticity and
walking performance was investigated by Ng and
Hui-Chan.15 The strength of the affected ankle
plantar flexors correlated well with the TUG
scores. A significant, negative relationship was
observed between the distance covered during the
6 Minute Walk Test (6MWT) and the TUG scores.
Flansbjer et al13 showed the TUG to correlate
positively with stair ascending and descending,
whereas a negative association was found between
fast and comfortable gait speed and the 6MWT.
Jnsdttir and Cattaneo5 found good convergent
validity by comparing the TUG and the Dynamic
Gait Index (DGI) in patients at least 3 months after
stroke. A good correlation was found between the
Community Balance and Mobility Scale (CB&M)
and the TUG and between the Berg Balance Scale
(BBS) and the TUG.14

Reliability

Three studies investigated the test-retest

reliability of the TUG13,15,23 and revealed excellent
intrarater and interrater reliability, with ICC values
greater than 0.95, as shown in Table 2.

Responsiveness and sensitivity to change

Four studies investigated the responsiveness

of the TUG12 and whether it is sensitive to
change.13,14,23 The TUG was found to be responsive

Prospective
longitudinal
study
n = 98 stroke
patients
n = 110 control
subjects
Validity study
N = 44

Simpson et
al, 201122
Canada

Hollands et
al, 201018
UK

Descriptive
study
n = 18
(50% stroke
patients with
fall history)
n = 18 control
subjects

Prospective
study
N = 96

Persson et
al, 201121
Sweden/
Norway

Knorr et al,
201014
Canada

Reliability study
N = 16

Design and
sample

Community dwelling

Community
dwelling; 3 months
after stroke;
follow-up after 5
months

Inpatient
rehabilitation wards;
discharged to their
own homes

Stroke unit;
follow-up at 3, 6, 12
months after stroke

Living at home

Setting follow-up

Differences between
stroke patients and
control subjects in
turning during the
TUG

Convergent validity,
sensitivity to change,
ceiling effect of CB&M
vs TUG, BBS, CMSA

Patients rates, location,

and predictors of falls
compared with those
of matched control
subjects (BBS, TUG,
6MWT, ABC)

Likelihood of clinical
tests identifying risk of
falling (TUG, 10MWT,
BBS, M-MAS, SPASS)

Reliability of 7
performance-based
tests (walking speed,
stair climbing, TUG)

Objective of the
study
Mean of 3
trials: 2 on the
first day, a third
7 days later.
IntraCC = 0.85;
InterCC = 0.96

Reliability

Significant convergent
abilities between
CB&M and TUG; P <
.001, r = -.75.

Validity

Study characteristics and clinimetric properties of the Timed Up and Go test

Faria et al,
201223
Brazil

Study

Table 2.

TUG baseline:
16.7 17.1;
follow-up: 13.7
16.0; SRM
0.34; P < .010

SEM of 3 trials,
1.49;
SEM%, 10.94

Responsiveness

Results

Although
participants
with stroke and
fall history took
significantly longer
to turn (mean 4.4
1.7 seconds)
than age-matched
control subjects (2.5
0.6 seconds), no
kinematic differences
were found in
turning performance.

Optimal cutoff score

for fall prediction:
15 seconds.
AUC 0.70 (95% CI,
0.60-0.81); PPV
58%;
NPV 63%
(regression
coefficient) -0.047;
IRR 0.955; 95% CI,
0.914-0.997 (P=0
.038).
Cutoff 15 seconds

Prediction of falls

Comments

(Continued)

8.1 seconds was

selected as maximum
score; 22.7% scored
max, and 36.4%
scored max in
follow-up

Patients unable to
perform the test are
included as median.
Performed during the
first week after stroke;
all tests had moderate
predictive value.
Decrease in TUG
time follow-up was
associated with
increased risk of
falling.

Only 1 trial was

necessary to provide
consistent and reliable
results.

Timed Up and Go: Clinimetric Properties After Stroke

201

Day hospital;
patients 3 months
after stroke

Community
dwelling

Descriptive
cohort study
N = 25

Cross-sectional
study
N = 50

Descriptive
study
N = 196

Jnsdttir
and Cattaneo,
20075
Italy
Belgen et al,
200619
USA

Andersson et
al, 200620
Sweden

Stroke unit;
follow-up after 6 or
12 months

Geriatric day hospital

Cross-sectional
study
N = 25

Heung and
Ng, 200916
Hong Kong

Community
dwelling;
2-155 months after
stroke

Setting follow-up

Descriptive
study
n = 22
n = 22 control
subjects

Design and
sample

(Continued)

Faria Coelho
de Morais et
al, 200917
Brazil

Study

Table 2.

Prediction of falls

Frequency of falls;
related to self-efficacy
and mobility measures

Reliability and validity

DGI; correlation with
TUG, BBS, and ABC

Differences in time
score of the TUG with
various chair heights
(65%, 90%, 115%
of each subjects leg
length)

Differences in time
score of the TUG
between stroke
patients and control
subjects when side of
turning was compared

Objective of the
study
Reliability

TUG: good negative

correlation with DGI;
r = -.77

Validity

Responsiveness

Results

Score TUG:
nonfallers, 16.0 7.6;
one-time and multiple
fallers, 19.9 13.9; no
significant difference
PPV 59, NPV 72;
cutoff score TUG
14 seconds

Prediction of falls

Comments

(Continued)

AUC .46, indicating

prediction near change;
strength partially
explained difference in
TUG score
TUG is a possible
choice for predicting
falls; 66% could
perform it.

No significant
difference in scores
when turning to
affected or unaffected
side.
Mean difference
between 2 trials: 2.45
(95% CI, 1.10-3.81)
Turning to affected side
significantly faster (P
< .001).
Lowest score when seat
height was 115% of leg
length (27.45 vs 32.25
seat height 65%).

202
TOPICS IN STROKE REHABILITATION/MAY-JUNE 2014

Test-retest
reliability study
N = 50

Cross-sectional
study
n = 11
n = 10 control
subjects

Prospective
cohort study
N = 50
divided into
slow, moderate,
fast walking
(5MWT speed
scores)

Flansbjer et
al, 200513
Sweden

Ng and HuiChan, 200515

Hong Kong

Salbach et al,
200112
Canada

Acute care hospital;

assessments 8 and 28
days after stroke

Community dwelling

Setting follow-up
Intra-rater (between
occasions, same day)
reliability study of
performance-based
tests after stroke
(walking speed, stair
climbing, TUG)
Test-retest reliability
of TUG performed on
different days 1 week
apart; association
with plantar flexion,
strength, gait speed
Responsiveness
of gait speed measures
5MWT and 10MWT
comfortable and
maximum speed, BBS,
BI, TUG, STREAM

Objective of the
study

Intra- and
interrater
reliability
Both ICCs >
.95

All the tests

highly reliable
IntraCC =
0.94-0.99;
TUG 0.96

Reliability
Significant correlation
found between
the TUG and CGS
(-0.86), FGS (-0.91),
SCas (0.86), SCde
(0.90), and 6MWT
(-0.92)
Strength plantar
flexion (P < .01), gait
parameters (P < .05),
walking endurance
(P < .01), significant
association

Validity

Results

SRMs for:
BBS 1.04;
BI .99;
STREAM .89;
TUG .73

SEM 1.14;
SEM% 8.2;
SRD% 23%;
95% SRD, 3.752.59

Responsiveness

Prediction of falls

In the fast and

moderate group,
most responsive were
5MWT and the TUG.
Ceiling effect: 32%;
2nd assessment:
8.5 seconds

TUG scores able to

differentiate between
patients and healthy
elderly

Reasonably small
improvements are
sufficient to detect real
changes.

Comments

Note: ABC = Activities-specific Balance Confidence Scale; AUC = area under the curve; BBS = Berg Balance Scale; BI = Barthel Index; CB&M= Community Balance and Mobility Scale; CI =
confidence interval; CMSA= Chedoke McMaster Stroke Assessment,; DGI = Dynamic Gait Index; FGS = Fast Gait Speed; 5-10MWT= 5 and 10 Meter Walking Test; GCS = Comfortable Gait
Speed; InterCC = interclass correlation coefficient: IntraCC = intraclass correlation coefficient; IRR = incidence rate ratio; MMAS = Modified Motor Assessment Scale; NPV = negative predictive
value; PVV = positive predictive value; SCas = stair climbing ascending; SCde = stair climbing descending; SEM = standard error of measurement; SEM% = limit for the smallest changes that
indicate real improvement; 6MWT = 6 Minute Walk Test; SPASS= Swedish Postural Assessment Scale for Stroke Patients; SRD = smallest real difference; SRM = standardized response mean
(= mean change divided by standard deviation of change); STREAM= Stroke Rehabilitation Assessment of Movement.

Design and
sample

(Continued)

Study

Table 2.

Timed Up and Go: Clinimetric Properties After Stroke

203

n = 80 stroke patients
living at home
n = 90 control subjects

Heung and Ng,

200916
Hong Kong

Faria Coelho
de Morais et al,
2009 17
Brazil

Hollands et al,
201018
UK

N = 25
patients with subacute
stroke

n = 22 control subjects

n = 18 control subjects
n = 22
2-155 months after
stroke, community
dwelling

Chair height 65%,

90%, 115% of leg
length (lateral knee
joint to floor); armrests
not mentioned

Armchair. height 100%

of subjects leg length

Seat height adjusted

with knees at 90,
forearms resting on
armrests

Armchair,
no specifications
reported

Not reported

N = 96
Assessment first week
after stroke

Persson et al,
201121
Sweden/
Norway
Simpson et al,
201122
Canada

N = 44
3 months
after stroke;
follow-up
after 5 months;
community dwelling
n = 18
stroke patients with and
without falls,
community dwelling

Height 100% leg

length; depth 45 cm;
width 49 cm; backrest
90 to horizontal;
armrest 20 cm
Standardized armchair;
height not mentioned;
sit with back supported

N = 16
Stroke patients living at
home

Faria et al,
201223
Brazil

Knorr et al,
201014
Canada

Chair

Sample and setting

Sit leaning against

chair back, hands
on thighs; turning
(line on the floor)
to affected and
nonaffected sides

At the word go,

walk at self-selected,
comfortable speed;
turn around, walk
back, and sit down

Turning at mark on
the floor; walking
as fast and safely as
possible

Not reported

Sit comfortably; on
command, go; selfselected speed; turn
around; walk back;
sit down
Walk as fast and
safely as possible;
turning point
marked by tape
Stand up, walk 3 m,
turn around, walk
back, and sit down

Instruction

2 trials to become
familiar; first
trial, turn toward
preferred side;
second, turn
toward opposite
side
2 trials to become
familiar (seat
65%); 3 trials
timed and 1-min
rest between trials

20x, 10x turning

toward the paretic
and nonparetic
sides

Not reported

First time, 2
assessments;
second
assessment after
7 days
Not reported
how many times
the test was
performed
Not reported

Procedure

Inclusion when
patients could walk
unassisted

Inclusion only
when patient could
perform the test.

All persons could

perform the test.

n = 2 at baseline;
in follow-up, all
patients could
perform the test.

All persons could

perform the test.

n = 28

All patients could

perform the test.

Patients not able

to perform test

The Timed Up and Go Test: Procedures and scores as described in the studies included

Study

Table 3.

Walking with or without

a walking aid;
n = 22 using a walking
aid

Walking with or without

assistive device

Walking without
assistance or walking aid;
number of patients
walking with device not
reported

Walking with or without

devices; number of
patients walking with
device not mentioned
Not reported

Support allowed and

recorded, as well as the
type of shoes

Preferred assistive devices

allowed

Use of walking aids

(Continued)

Lowest score seat height

115% of leg length:
27.45 seconds;
seat height 65%:
32.25 seconds

Healthy control subjects

mean time: 9.3 seconds

Control: 12-27.5 seconds

Turning time: 4.4 vs 2.5
seconds
Stroke patients
mean time: 28.9 seconds

TUG mean time: 18-49

seconds

Control: 8.2 1.8 seconds

Baseline:
16.7 17.1 seconds
Follow-up:
13.7 16.0 seconds

Mean TUG time: 20

14.3 seconds

Mean of first 2 trials:

12.65 3.76 seconds
Score: 8-59 seconds
Median: 15 seconds

Mean of 3 trials: 12.37

3.54 seconds

Scores

204
TOPICS IN STROKE REHABILITATION/MAY-JUNE 2014

N = 159
patients treated in a
stroke unit;
follow-up after 12
months

N = 50
6-46 months
after stroke;
community
dwelling;
able to walk

n = 11
5.6 3.3 years
after stroke

n = 10 control subjects
N = 50
patients with acute
stroke;
follow-up after 4 weeks

Andersson et
al, 200620
Sweden

Flansbjer et al,
200513
Sweden

Ng and HuiChan, 200515

Hong Kong

Salbach et al,
200112
Canada

Armchair;
no specification
reported

Chair with armrests;

height not reported.

Height 44 cm;
depth 45 cm;
width 49 cm;
armrest height
64 cm

No specifics
reported

Back against the

chair; on the word
go, stand up,
walk at comfortable
speed, pass the
3-m mark, walk
back, sit down.
No specifics
reported

Walk 3 m as fast as
possible, cross a line
on the floor, turn,
walk back, and sit
down

Rise, walk 3 m at
usual pace, turn
around, walk back,
and sit down

Armchair
height not reported

Armchair
height not reported

Not reported

No specification
reported

N = 25
3 months
after stroke;
able to walk
N = 50
62.2 62.1 months after
stroke

Jnsdttir
and Cattaneo,
20075
Italy
Belgen et al,
200619
USA

Instruction

Chair

Sample and setting

(Continued)

Study

Table 3.

Examiner stands
to the side while
timing

5-11 days after

stroke, TUG
performed twice;
second time,
carrying a glass of
water
1 trial to become
familiar; 1-min
rest; performed
twice, with 1min rest between;
mean of the 2
tests recorded
2 assessments on
different days in
a week

TUG performed
once
(TUG score
related to falls)

Not reported

Procedure

n = 10 not able to
perform the test at
baseline;
n = 3 after 4 weeks.

Inclusion only
when patient could
walk and perform
the test.

Inclusion only
when patient could
walk and perform
the test.

n = 105 (66%) were

able to perform the
test at baseline.

Inclusion only
when patient could
walk and perform
the test
Inclusion only
when patient could
walk and perform
the test

Patients not able

to perform test

Use of walking device or

physical assistance was
permitted.

Walking with or without

walking aid;
n = 9 used a walking aid.

n = 32 no walking aid
n = 7 used ankle- foot
orthesis

Not reported

Walking with or without

walking aid;
n = 27 with no assistive
device

Walking with or without

walking aid

Use of walking aids

Healthy subjects:
9.1 1.6 seconds
Stroke patients:
22.6 8.6 seconds
P < .001
26.2 17.3 seconds
Follow-up: 19.6 17.5
seconds

First session mean: 14.3

5.2 seconds
(range, 7.5-25.7 seconds)
Second session mean:
13.7 5.3 seconds
(range, 6.7-27.7 seconds)

Nonfallers 16.0 7.6

seconds
One-time and multiple
fallers 19.9 13.9
seconds
No significant difference
between groups
Absolute scores not
reported

24.5 11.8 seconds

(range, 7.1-55.5 seconds)

Scores

Timed Up and Go: Clinimetric Properties After Stroke

205

206

TOPICS IN STROKE REHABILITATION/MAY-JUNE 2014

in moderate- and fast-walking patients with

stroke.12 The TUG was sufficiently sensitive to
detect small changes in patients with stroke,
as the percentage of the standard error of the
measurements (SEM%) was 10.9%. An SEM%
<15% indicates that the test is sensitive and can
be used to detect small changes in patients with
stroke.23 The TUG was also able detect small changes
in patients with mild chronic stroke,13 whereas
in another study, the effect size (ES) was small.12
However, 2 studies demonstrated a ceiling effect for
the TUG,11,13 which was defined as more than 20%
of the participants achieving the maximum score.24
In one study, 22% of the participants performed the
test in 8.1 seconds or less.14 In another study, 8%
of the patients performed the test in 8.5 seconds or
less in the first evaluation, and 32% of the patients
performed the test in 8.5 seconds or less in the
second evaluation, which was performed after 4
weeks, demonstrating a ceiling effect.12

the TUG, but only half of this group had a fall

history.18 The rate of falls for patients with stroke
was 1.77 times that of individuals in the control
group. The baseline mean TUG score for the stroke
group was 20 seconds, compared with 8.2 seconds
for the control group.22
Procedures

As shown in Table 3, the studies provide poor

and varying descriptions of the procedures used
to apply the TUG. Many studies do not report
the type or height of the chair, the marker used
to indicate the turning point, or whether footwear
was worn, the instructions given, or a walking
aid was allowed. Additionally, normative scores
for patients with stroke are lacking. The mean
scores in the studies including only patients with
stroke were 14.3 seconds,13 22.6 seconds,15 and
27.4 to 34.9 seconds, depending on chair height
and turning side,16 and 25.8 to 28.96 seconds,
depending on the turning side.17

Predictability for falls

Five studies investigated the TUG in relation to

falls after stroke. Of these, 3 studies investigated
whether the TUG could predict falls after
stroke,19-21 and 2 studies examined whether the
TUG could differentiate between fallers and
nonfallers after stroke.18,22 The TUG could predict
the risk of falling during the first year after stroke
if performed in the first week after the stroke
occurred, with an optimal cutoff point of 15
seconds. Another study showed that the TUG
score of fallers differed significantly from that of
nonfallers using a cutoff score of 14 seconds.20
The fallers took longer to finish the TUG than
the nonfallers.21 Another study, however, showed
that the TUG could not predict falls in the chronic
phase after stroke, as no significant difference
in TUG scores was found between groups.19
The ability of the TUG to differentiate between
fallers and nonfallers after stroke while turning
was investigated in 2 studies.18,22 Hollands et al18
found the time required for patients to turn to be
significantly longer for fallers than for nonfallers.
It took patients with stroke who had a fall history
significantly longer to turn compared with agematched control subjects. All the patients with
stroke required more than 14 seconds to perform

Walking aids

In the original instructions of the TUG, the

patient is allowed to walk with or without a
walking aid.4 Only 2 studies failed to indicate
whether patients used a walking device, 14,15
and 4 studies indicated the number of patients
walking with a device.13,15,16,19 None of the authors
described the difference in the time required for
patients walking with a walking aid and those
walking without a walking aid.
Turning to the affected or unaffected side

Two studies explored whether the TUG score

differed when the patient turned to the affected or
unaffected side or when chairs of different heights
were used.16,17 One study investigated the TUG
using 6 conditions, namely, turning to the affected
or unaffected side and using chairs of different
heights.16 The mean TUG scores obtained when
using chairs of different heights were significantly
different. Patients with stroke completed the TUG
more quickly when turning to the affected side,
with no significant interaction between chair
height and turning direction. The lowest TUG
scores were recorded with a chair height equal

Timed Up and Go: Clinimetric Properties After Stroke

to 115% of the subjects leg length (mean, 27.4

seconds).16 In the second study, when the TUG
scores of patients with stroke were compared with
those of a healthy control group, no significant
difference was found in the length of time required
for turning to either the affected or unaffected
side.17
Discussion
This review showed that the TUG has excellent
reliability when used for patients with stroke,
confirming the findings of earlier studies.4,13,15,23
The TUG showed good convergent validity, as
the scores correlated well with other measures of
functional mobility. The scores of the investigated
tests, including the 6MWT, Comfortable and
Fast Gait Speed test, and stair climbing,5 were
highly related to the scores on the TUG.13-15 Four
studies showed the TUG to be sensitive to small
changes in the basic functional mobility of patients
with stroke.12-14,23 This finding is important for
professionals in various health care settings who
need to be able to evaluate the often slow and
small progress made by patients in basic functional
mobility after stroke. In 2 of the studies, a ceiling
effect for the TUG was reported,12,14 and this
was most evident for patients dwelling in the
community with moderate and mild deficits.25
Although the ceiling effect indicates that the TUG
may have less value for people who have reached
good functional ability, the TUG is still a good
instrument for detecting functional progress in
patients with stroke.
The TUG has been used as a tool to predict the
fall risk of patients with stroke. In the literature
on elderly people without stroke, different cutoff
points are reported, which vary from 12.4725
to 15 seconds.24 In one study of patients with
stroke, in which a cutoff point of 14 seconds was
used, a difference was found between fallers and
nonfallers, with a positive predictive value (PPV)
of 59% and a negative predictive value (NPV)
of 72%.20 Unexpectedly, a lower TUG score was
associated with an increase in falls in patients
who were recently discharged to their homes after
stroke,22 indicating that the fall risk increases as
functional mobility improves. This finding may
be explained by the fact that when patients have

207

recently been discharged to their homes, they

experiment with walking in surroundings where
they have not practiced walking after the stroke,
which increases the risk of falling.
Although one study indicated that the TUG
best predicts falls when a cutoff score of 15
seconds is used (PPV 58% and NPV 63%),21
another study indicated a prediction near chance
(receiver operating curve, .50). 19 The cutoff
points for predicting the fall risk after stroke are
inconclusive, and the PPVs are low. The TUG can
be used as an indicator of fall risk but must be used
in combination with other measures, never as the
only predictive tool.26
The studies included provided poor descriptions
of how the TUG was performed, and when
descriptions were provided, they differed
considerably. Procedural differences affect the
scores on the TUG. 27, 28 Interestingly, Heung
and Ng16 found that the height of the chair and
the turning direction influence the TUG scores.
Patients were able to turn to the affected side more
quickly than to the unaffected side. The number
of steps used in the turning procedure might have
influenced the TUG time, because turning to the
unaffected side requires more steps. Faria Coelho
de Marais et al,17 however, found no difference in
the TUG scores when subjects were turning. It is
important to have a standardized procedure for
applying the TUG. Professionals need to be able
to compare scores when evaluating a patients
progress. Using an observation sheet to record the
circumstances under which the TUG is applied,
including the use of a walking aid and the turning
side, would be ideal. Another important finding
concerns the lack of normative time scores on the
TUG for patients with stroke compared with those
of relatively healthy elderly people.29,30
In 2 prior meta-analyses that reviewed
instruments for assessing walking balance after
stroke, including the TUG, the authors concluded
that the TUG addresses relatively few aspects of
balance.31,32 Outcome measures with multiple
tasks better reflect the skills necessary for walking
in the community, whereas the TUG was judged
as a single task measure.32 We, however, judge
the TUG to be a measurement of basic mobility
properties, because the TUG was not originally
developed to measure walking balance. 4 In

208

TOPICS IN STROKE REHABILITATION/MAY-JUNE 2014

daily life, standing up from a chair, walking a

few meters, turning around, and sitting down
are essential basic abilities required to function
without help when living at home.
With regard to the clinical relevance of the
TUG score in relation to safely walking at home,
the categorization of Posiadlo and Richardson4
remains useful. Patients can be divided into 3
categories based on their scores on the TUG: (1)
patients who perform the test in less than 20
seconds tend to be independent mobile and
have a gait speed of at least 0.5 meters per second,
which is the minimum velocity required to walk
safely in the community; (2) patients who perform
the test in more than 30 seconds generally need
assistance in basic activities; and (3) patients who
perform the test between 20 and 30 seconds need
further assessment to define their functional level.4
Functional performance tests, such as the
TUG, can detect short-term functional changes
after stroke. However, there is the danger of
measurement error.33 Two measurements must
be performed to detect change, and no single
change measured is necessarily a true change.
The error can be due to a learning effect, fatigue,
or tester inconsistency. Any performance change
must be greater than the error. The 4 studies that
investigated the responsiveness of the TUG show
good responsiveness to change.12-14,23
Furthermore, patients scores on the TUG differ
depending on the use of a walking aid, as patients
may use no cane, a cane, or a walker. In only
one study was a time difference of more than 2
seconds described in patients (not patients with
stroke) walking with or without a cane.27 Three
of the included studies15-17 included reports of
the number of patients who used a walking aid
during the test, but these studies did not indicate
the difference in scores. Notably, one of the
studies that investigated various instruments for
identifying patients at risk of falling also included
patients who were unable to walk.21
The feasibility of the TUG is another important
factor. No study investigated the use of the TUG by
professionals other than physical therapists. The
TUG is a feasible instrument for other professionals
as well, including nurses, because nurses need
to establish the basic mobility skills of patients
with stroke who can walk, judge whether their

walking is safe, and monitor patients progress.

However, before structural implementation in
clinical practice, it is highly important to provide
these professionals with appropriate education
and training on how to use the TUG. Standardized
procedures need to be used in applying the TUG,
and the inter- and intrarater reliability of the TUG
must be tested before the TUG is applied in the
daily management of patients with stroke.
Limitations

Although a comprehensive and systematic

search of the literature was conducted, it is
possible that relevant studies may have been
missed. The heterogeneity of the designs used, the
clinimetric properties measured, and the patients
included made it difficult to arrive at sound
conclusions, which may limit the interpretability
of the clinimetric properties investigated in these
studies. However, a thorough and transparent
procedure was followed for the inclusion and
selection of studies, as well as for the judgment of
methodological quality, which was performed by 2
independent reviewers.
Conclusion
The TUG includes many aspects of basic
mobility skills. The findings of this review show
that the TUG has excellent intra- and interrater
reliability and good construct and convergent
validity and is sensitive enough to detect small
changes in basic functional mobility after stroke.
However, the results are inconclusive regarding
the ability of the TUG to predict patient falls after
stroke. There were substantial differences in the
procedures and instructions used to apply the
TUG. Based on the findings of this review, the
TUG can be recommended for use by various
professionals in stroke care to monitor the basic
functional mobility of patients with stroke who are
able to walk, assuming that the professionals are
properly trained in how to apply the TUG and use
standardized procedures.
Acknowledgments
The authors report no conflicts of interest.

Timed Up and Go: Clinimetric Properties After Stroke

209

REFERENCES
1. Go AS, Mozaffarian D, Roger VL, et al. Heart
disease and stroke statistics2013 update: A
report from the American Heart Association
Circulation 2013;127:e6-e245. doi: 10.1161/
CIR.0b013e31828124ad.
2. Czernuszenko A. Risk factors for falls in post-stroke
patients treated in a neurorehabilitation ward.
Neurol Neurochir Pol. 2007;1:2835.
3. L a n g h o r n e P, B e r n h a r d J , K w a k k e l G .
Stroke rehabilitation: Stroke care 2. Lancet.
2011;377:16931702
4. Posiadlo D, Richardson S. The timed Up & Go:
A test of basic functional ability for frail elderly
persons. J Am Geriatr Soc. 1991;39:142148.
5. Jnsdttir J, Cattaneo D. Reliability and validity
of the dynamic gait index in persons with
chronic stroke. Arch Phys Med Rehabil. 2007;88:
14101415.
6. Brooks D, Davis AM, Naglie G. Validity of 3
physical performance measures in inpatient
geriatric rehabilitation. Arch Phys Med Rehabil.
2006;87:105110.
7. VanIersel MB, Munneke M, Esselink RA, Benraad
CE, Olde Rikkert MG. Gait velocity and the Timed
Up and Go Test were sensitive to changes in
mobility in frail elderly patients. J Clin Epidemiol.
2008;61:186191.
8. Demers L, Desrosiers J, Nikolova R, Robichaud L,
Bravo G. Responsiveness of mobility, daily living
and instrumental activities of daily living outcome
measures for geriatric rehabilitation. Arch Phys Med
Rehabil. 2010;91:233240.
9. Moher D, Liberati A, Tetzlaff J, Altman DG,
PRISMA Group. Preferred reporting items for
systematic reviews and meta-analyses: The PRISMA
statement. BMJ. 2009;339:332336.
10. Roy JS, Desmeulin F, MacDermid JC. Psychometric
properties of presenteeism scales for musculoskeletal
disorders: A systematic review. J Rehabil Med.
2011;43:2331.
11. Roy JS, MacDermid JC, Woodhouse LJ. Measuring
shoulder function: A systematic review of four
questionnaires. Arthritis Rheum. 2009;61:623632.
12. Salbach NM, Mayo NE, Higgins J, Ahmed S, Finch
LE, Richards CL. Responsiveness and predictability
of gait speed and other disability measures in
acute stroke. Arch Phys Med Rehabil. 2001;82:
12041212.
13. Flansbjer UB, Holmback AM, Downham D, Patten
C, Lexell J. Reliability of gait performance tests in
men and women with hemiparesis after stroke.
J Rehabil Med. 2005;37:7582.
14. Knorr S, Brouwer B, Garland J. Validity of the
community balance and mobility scale in
community-dwelling persons after stroke. Arch
Phys Med Rehabil. 2010;91:890896.
15. Ng SS, Hui-Chan CW. The Timed Up and Go
test: Its reliability and association with lowerlimb impairments and locomotor capacities in
people with chronic stroke. Arch Phys Med Rehabil.
2005;86:16411647.

16. Heung THM, Ng SSM. Effect of seat height and

turning direction on the timed up and go test
scores of people after stroke. J Rehabil Med.
2009;41:719722.
17. Faria Coelho de Marais CD, Teixeira-Salmela LF,
Nadeau S. Effects of the direction of turning on
the Timed Up and Go Test with stroke subjects. Top
Stroke Rehabil. 2009;16:196206.
18. Hollands KL, Hollands MA, Zietz D, Wing AM,
Wright C, van Vliet P. Kinematics of turning 180
during the Timed Up and Go in stroke survivors
with and without fall history. Neurorehabil Neural
Repair. 2010;24:358367.
19. Belgen B, Beninato M, Sullivan PE, Narielwalla
K. The association of balance capacity and falls
self-efficacy with history of falling in communitydwelling people with chronic stroke. Arch Phys Med
Rehabil. 2006;87:554561.
20. Andersson G, Kamwendo K, Seiger , Appelros
P. How to identify potential fallers in a stroke unit:
Validity indexes of four test methods. J Rehabil
Med. 2006;38:186191.
21. Persson CU, Hansson PO, Sunnerhagen KS.
Clinical tests performed in acute stroke identify
the risk of falling during the first year: Postural
stroke study in Gothenburg. J Rehabil Med.
2011;43:348353.
22. Simpson LA, Miler WC, Eng JJ. Effect of stroke on
fall rate, location and predictors: A prospective
comparison of older adults with and without
stroke. PloS One. 2011;6 (4):e19431.
23. Faria CD, Teixeria-Salmelal LF, Neto MG,
Rodriques-de Paula F. Performance based tests in
subjects with stroke: Outcome scores reliability
and measurement errors. Clin Rehabil. 2012;26(5):
260269.
24. Nordin E, Lindelof N, Rosendahl E, Jensen J,
Lundin-Olson L. Prognostic validity of the Timed
Up and Go test, a modified Get up and Go
test, staffs global judgment and fall history in
evaluating fall risk in residential care facilities. Age
Ageing. 2008;37:442448.
25. Shumway-Cook A, Brauer S, Woollacott M.
Predicting the probability for falls in community
dwelling older adults using the Timed Up & Go
test. Phys Ther. 2000;80:896903.
26. R e n s i n k M , S c h u u r m a n s M , L i n d e m a n ,
Hafsteinsdttir T. Vallen: Incidentie en risicofactoren
bij patinten met een beroerte. Een systematisch
literatuur onderzoek. Tijdschr Gerontol Geriatr.
2009;40:156168.
27. Kalula SZ, Swingler GH, Sayer AA, Badri M, Ferreira
M. Does chair type influence outcome in the
Timed Up and Go test in older persons ? J Nutr
Health Aging. 2010;14(4):319323.
28. Bergmann JH, Alexiou C, Smith IC. Procedural
differences directly affect timed up and go times. J
Am Geriatric Soc. 2009;57:21682169.
29. Bohannon RW. Reference values for the Timed Up
and Go Test: A descriptive meta-analysis. J Geriatr
Phys Ther. 2006;29:6468.

210

TOPICS IN STROKE REHABILITATION/MAY-JUNE 2014

30. Pondal M, del Ser T. Normative data and

determinants for the timed up and go test in
a population-based sample of elderly individuals
without gait disturbances. J Geriatr Phys Ther.
2008;31:5763.
31. Oliviera C, Medeiros IR, Frota NA, et al. Balance
control in hemiparetic stroke patients: Main tools for
evaluation. J Rehabil Res Dev. 2008;45:12151226.

32. Pollock CL, Eng JJ, Garland SJ. Clinical

measurement of walking balance in people
post stroke: A systematic review. Clin Rehabil.
2010;25:693708.
33. Streiner DL, Norman GR. Health Measurements
Scales: A Practical Guide to Their Development and
Use. 3rd ed. New York: Oxford University Press;
2003.