Gait speed and risk assessment for falls among men aged 80 years and older a prospective cohort study in taiwan

G Model

EURGER-511; No. of Pages 5
European Geriatric Medicine xxx (2014) xxx–xxx

Available online at

ScienceDirect
www.sciencedirect.com

Research paper

Gait speed and risk assessment for falls among men aged 80 years and
older: A prospective cohort study in Taiwan
C.-K. Liang a,b,d, M.-Y. Chou a,d,e, L.-N. Peng c,d, M.-C. Liao a,f, C.-L. Chu a,g, Y.-T. Lin a,b,d,*,
L.-K. Chen c,d,**
a

Geriatric Medicine Center, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan
Division of Neurology, Department of Internal Medicine, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan
Center for Geriatrics and Gerontology, Taipei Veterans General Hospital, Taipei, Taiwan
d
Aging and Health Research Center, National Yang Ming University, Taipei, Taiwan
e
Department of Family Medicine, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan
f
Department of Emergency Medicine, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan
g
Department of Psychiatry, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan
b
c

A R T I C L E I N F O

A B S T R A C T

Article history:
Received 20 April 2014
Accepted 6 June 2014
Available online xxx

Purpose: To evaluate the effectiveness of adding gait speed to the history of falls in predicting falls among
men aged 80 years and older in Taiwan.
Methods: This prospective cohort study recruited 230 ambulatory men aged 80 years and older in 2012
and followed for 12 months. In addition to demographic characteristics and history of falls, a
comprehensive geriatric assessment was performed for all study subjects. Gait speed was obtained by
the 6-m walk and three different cut-offs (< 0.5, 0.8 and < 1.0 m/s) were tested in improving the
ability of predicting subsequent falls by using history of falls.
Results: Among all subjects (mean age: 85.5 Æ 4.0 years), 26.1% (60/230) reported falls during follow-up
period. Univariate analysis showed that polypharmacy, urinary incontinence, history of falls, pain, poorer
baseline physical function, depressive mood, and gait speed < 0.5 m/s were associated with falls. Logistic
regression showed that history of falls (OR: 4.255, 95% CI 2.089–8.667; P < 0.001), pain (OR: 2.674, 95% CI
1.332–5.369; P = 0.006), older age (OR: 1.128, 95% CI 1.031–1.234; P = 0.008), and slow gait speed (OR: 2.964,
95% CI 1.394–6.300; P = 0.005) were all independent risk factors for falls. Fast gait speed (defined as ! 1 m/s)
was a protective factor for falls, even among subjects with history of falls, but slow gait speed (defined
as < 0.5 m/s) was an independent risk factor even among subjects without history of falls.
Conclusions: Combined history of falls and gait speed is a simple and effective tool in risk assessment of
falls among older old population.
ß 2014 Published by Elsevier Masson SAS.

Keywords:

Gait speed
Fall
Oldest old
Men
History of falls

1. Introduction
Taiwan has become an aging society in 1993 and is estimated to
become an aged society by 2017, which makes Taiwan the fastest
aging country in the world [1]. Population aging may cause various

* Corresponding author. Geriatric Medicine Center and Division of Neurology,
Kaohsiung Veterans General Hospital, Taiwan; No. 386, Ta-Chung 1st RD, Zuoying
District, Kaohsiung 813, Taiwan. Tel.: +886 7 342 2121x2091; fax: +886 7 348 1478.
** Corresponding author. Center for Geriatrics and Gerontology, Taipei Veterans
General Hospital, No. 201, Sec 2, Shih-Pai Road, Taipei, Taiwan. Tel.: +886 2 2875 7830;
fax: +886 2 2875 7711.
E-mail addresses: (C.-K. Liang),
(M.-Y. Chou), (L.-N. Peng),
(M.-C. Liao), (C.-L. Chu), (Y.-T. Lin),
(L.-K. Chen).

challenges to the health care systems, and falls have been associated
with strong risks to the health of older people. Generally speaking,
nearly a third of elderly people may experience falls every year, and
more than 50% of these falls occurred during certain form of
locomotion [2]. Falls are the most common cause of injury-related
deaths among people aged 75 and older, which is the same as in nonfatal injuries among females aged 85 years and older [3]. Falls are also
a serious public health issue that are highly associated with morbidity
and mortality of older people [4], and a multifactorial approach is

considered the most effective strategy to prevent falls [5].
In fall prevention programs, screening the risk of falls is the first
and the most critical step to stop the vicious cycle. Screening the
history of previous falls is a quick, simple, and effective tool for the
first step of risk assessment, which was supported by both
the American Geriatrics Society (AGS) and the British Geriatrics

/>1878-7649/ß 2014 Published by Elsevier Masson SAS.

Please cite this article in press as: Liang C-K, et al. Gait speed and risk assessment for falls among men aged 80 years and older: A
prospective cohort study in Taiwan. Eur Geriatr Med (2014), />

G Model

EURGER-511; No. of Pages 5
2

C.-K. Liang et al. / European Geriatric Medicine xxx (2014) xxx–xxx

Society (BGS) [6]. History of fall is the strongest predictor for
subsequent falls [7–9], as well as the risk for fractures among
people aged 45 years and over [10]. In the AGS-BGS guideline, after
screening the history of falls, evaluating the gait/balance
disturbance was the second step in the risk assessment, and a
number of tests were recommended, such as Get up and Go test,
Timed Up and Go test, the Berg Balance Scale, and the
Performance-Oriented Mobility Assessment [6]. However, currently, no sufficient evidence supported using a specific test for
balance and gait disturbance to predict subsequent falls [6]. Among
all these tests, gait speed has been recognized as a simple screening
test for various adverse health outcomes of older people, such as

mobility disability, institutionalization, death, and cognitive
decline [11]. It has been reported that the gait speed 0.8 m/s
was associated with a higher risk of adverse health outcomes,
[11,12] and a gait speed slower than 1.0 m/s may increase the risk
of mortality [13,14]. However, studies focused on the effect of gait
speed in predicting falls among people aged over 80 years were
scarce. Moreover, some previous studies suggested to re-define the
cut-off of slower gait speed among oldest old population due to
their survival effect [15,16]. Although the history of falls is an
effective screening tool for falls, it does not completely reflect the
current health status, physical function and risk for falls of older
people. Therefore, the main aim of this study was to evaluate the
role of gait speed in history of previous falls to predict subsequent
falls among people aged 80 years and older in Taiwan.

as currently using > 4 prescription drugs for over 2 weeks),
depressive symptoms using the 15-item Chinese Geriatric
Depression Scale (GDS-15, a score of 5 and more was defined as
depressive) [18], nutritional status using the Mini Nutritional
Assessment-short form (MNA-SF, malnutrition was defined as the
MNA-SF scores of 11 and less) [19], cognitive function determined
by the Chinese version of the Mini-Mental State Examination
(MMSE, the scores less than 24 was defined as cognitive
impairment) [20], the instrumental activities of daily living (using
the Lawton-Brody Instrumental ADL scale, IADL) [21], and quality
of life (using European quality of life–five domains, EQ5D) [22].
2.2.3. Gait speed measurement
A timed 6-m walk was performed for all participants at their
usual walking speed with a static start throughout a 6-m distance
without deceleration [23,24], and the time consumed was taken

by a fixed study nurse with a stop watch (HS-70 W, Casio
computer co. LTC, Tokyo, Japan). The test allowed the subjects to
start with a cane or a walker as needed. Three different cut-offs for
slower gait speed (< 0.5 m/s [25], 0.8 m/s [11,12], and < 1 m/s
[13,14]) were tested to evaluate the effect in improving fall
prevention among the study subjects.

2. Methods

2.2.4. Definitions of falls
In this study, a fall was defined as an unintentional change in
position resulting in coming to rest on the ground or other lower
levels [26]. For all study subjects, the occurrence of falls was
carefully recorded during the 12-month follow-up period.

2.1. Participants and study design

2.3. Statistical analysis

This prospective observational cohort study invited all residents living in the Veterans Home, a retirement community, in
southern Taiwan in January of 2012. For those who participated in
the study, demographic data were collected and the comprehensive geriatric assessments were preformed to them twice a year
after the enrollment. Subjects with the following conditions were
excluded for study:

In this study, continuous variables in the text and tables were
expressed as means with standard deviation, and categorical data
were expressed as percentages. Comparisons between dichotomous and ordinal variables were done by using the Chi2 test or
Fisher’s exact test when appropriate, and comparisons between
continuous variables were done using the independent Student’s

t-test or Mann-Whiney U test when appropriate. Multivariate
logistic regression analysis was used to determine the independent
predictive factors for subsequent falls in the following year and
the candidate predictors with a P value < 0.2 in univariate
analysis were selected to enter the regression model. For the
interaction of history of falls and gait speed, we combined
the history of falls and slower gait speeds using different
definitions, i.e. < 0.5 m/s, 0.8 m/s or < 1 m/s. The predictive
effect was also analyzing by multivariate stepwise logistic
regression analysis after adjusting confounders.






unable to walk with a walking aid;
unable to communicate with research staff;
unable to obtain informed consent from participants;
subjects with their expected life expectancy shorter than
12 months.

A total of 278 people aged 80 years and older were screened,
and 7 of them were excluded (5 people were unable to walk, 2
person with incomplete fall history) for study. Among eligible
study subjects (n = 271), 41 people did not complete the 6-m walk
test, so, only a total of 230 residents were enrolled in this study.
The whole study was approved by the Institutional Review Board
of Kaohsiung Veterans General Hospital.
2.2. Data collection

2.2.1. Demographic characteristics
Three well-trained research nurses interviewed all participants
to collect the demographic data, including age, smoking habit,
habitual alcohol use status, presence of pain, sleep problems,
urinary incontinence, medical history, co-morbidities by using
Charlson Comorbidity Index [17], and body mass index (BMI) were
obtained for each study subject.
2.2.2. Comprehensive Geriatric Assessment (CGA)
The research nurses performed CGA for all participants, which
included visual and hearing impairment, polypharmacy (defined

3. Results
3.1. Demographic characteristics and functional status
Overall, 230 subjects (mean age: 85.5 Æ 4.0 years, range:
80–101 years, all males) participated in this study and 27.4% of
them (63/230) reported falls in the previous year. Among them, 26.1%
(60/230) reported fall events during the follow-up period. Table 1
summarized the demographic characteristics and functional status of
the study participants. Approximately 40% of the study subjects had
sleeping problems, urinary incontinence, cognitive impairment, or
depressive symptoms. Those who developed falls in the follow-up
period had significantly slower gait speed than those who developed
no fall event (0.67 Æ 0.33 m/s vs 0.78 Æ 0.32 m/s, P = 0.021), lower
scores in EQ5D (61.1 Æ 22.9 vs 68.1 Æ 15.7, P = 0.039), and higher
prevalence of urinary incontinence (46.7% vs 27.6%, P = 0.007),
presence of pain (61.7% vs 41.2%, P = 0.001), and depressive
symptoms (43.3% vs 24.7%, P = 0.007).

Please cite this article in press as: Liang C-K, et al. Gait speed and risk assessment for falls among men aged 80 years and older: A
prospective cohort study in Taiwan. Eur Geriatr Med (2014), />

G Model

EURGER-511; No. of Pages 5
C.-K. Liang et al. / European Geriatric Medicine xxx (2014) xxx–xxx

3

Table 1
Comparisons of baseline characteristics among subjects with or without falls during the 12-month follow-up.
Total

Fall (+)

Fall (–)

Variables

% or mean Æ SD (n)

% or mean Æ SD (n)

% or mean Æ SD (n)

P value

Age
Current smoker (yes)
Current drinker (yes)
Sleep problems (yes)

Urine incontinence (yes)
Hx of fall in past 1 year
Presence of pain (yes)
BMI
CCI
Polypharmacy (yes)
Visual impairment (yes)
Hearing impairment (yes)
Baseline IADL
EQ5D VAS Scores
Cognitive impairment (yes)
Depressive symptoms (yes)
Risk of malnutrition (yes)
Gait speed (m/s)

85.5 Æ 4.0 (n = 230)
43/230 (18.7%)
65/230 (28.3%)
88/230 (38.3%)
75/230 (32.6%)
63/230 (27.4%)
107/230 (46.5%)
24.1 Æ 3.5 (n = 169)
1.01 Æ 1.48 (n = 230)
146/230 (63.5%)
183/230 (79.6%)
149/230 (64.8%)
6.9 Æ 1.3 (n = 230)
66.2 Æ 18.1 (n = 203)
88/230 (38.3%)

68/230 (29.6%)
51/230 (22.2%)
0.75 Æ 0.32 (n = 230)

86.1 Æ 4.2 (n = 60)
10/60 (16.7%)
15/60 (25.0%)
26/60 (43.3%)
28/60 (46.7%)
29/60 (48.3%)
37/60 (61.7%)
24.2 Æ 3.9 (n = 37)
1.08 Æ 1.52 (n = 60)
45/60 (75.0%)
48/60 (80.0%)
44/60 (73.3%)
6.6 Æ 1.6 (n = 60)
61.1 Æ 22.9 (n = 54)
24/60 (40.0%)
26/60 (43.3%)
12/60 (20.0%)
0.67 Æ 0.33 (n = 60)

85.3 Æ 3.9 (n = 170)
33/170 (19.4%)
50/170 (29.4%)
62/170 (36.5%)
47/170 (27.6%)
34/170 (20.0%)
70/170 (41.2%)

24.1 Æ 3.4 (n = 132)
0.99 Æ 1.46 (n = 170)
101/170 (59.4%)
135/170 (79.4%)
105/170 (61.8%)
7.1 Æ 1.1 (n = 170)
68.1 Æ 15.7 (n = 149)
64/170 (37.6%)
42/170 (24.7%)
39/170 (22.9%)
0.78 Æ 0.32 (n = 170)

0.110
0.639
0.514
0.347
0.007
< 0.001
0.001
0.850
0.669
0.031
0.923
0.107
0.038
0.039
0.747
0.007
0.637
0.021

BMI: Body mass index; CCI: Charson Comorbisity Index.

3.2. Independent risk factors for falls
All candidate predictors with the P value < 0.2 in the
univariate analysis were entered into the logistic regression
analysis, including age, polypharmacy, hearing impairment,
urinary incontinence, history of falls, EQ5D scores, presence of
pain, IADL, depressive symptoms, and gait speed. Results
showed that history of falls in the previous year (odds ratio
[OR]: 4.255, 95% confidence interval [CI]: 2.089–8.667;
P < 0.001), presence of pain (OR: 2.674, 95% CI: 1.332–5.369;
P = 0.006) and older age (OR: 1.128, 95% CI 1.031–1.234;
P = 0.008) were all independent risk factors for falls (Table 2).
Although the gait speed per se was not an independent
predictive factor for falls, but the slower gait speed (defined
by < 0.5 m/s) was independently associated with falls in the
subsequent year (OR: 2.964, 95% CI: 1.394–6.300; P = 0.005).
3.3. Synergic effect of history of falls and gait speed
Adjusted for age, polypharmacy, hearing impairment, urinary
incontinence, EQ5D scores, presence of pain, IADL, and depressive
symptoms, we evaluated the potentially synergic effect of history
of falls and slower gait speed in fall prediction (Table 3). A slower
gait speed did not significantly add predictive value on history of
falls if it was defined as 0.8 m/s (OR: 4.044, 95% CI: 1.504–
10.869, P = 0.006 for subjects with history of falls and slower gait
speed; OR: 4.290, 95% CI: 1.379–13.346, P = 0.012 for subjects with
history of falls and faster gait speed). Moreover, if slower gait speed
was defined as < 0.5 m/s, a strong synergic effect was identified

(OR: 3.308, 95% CI: 1.280–8.549, P = 0.014 for subjects with slow
gait speed without history of falls; OR: 4.423, 95% CI: 1.857–
10.535; P = 0.001 for subjects with history of falls and no slow gait
speed; OR: 10.920, 95% CI: 3.423–34.839, P < 0.001 for subjects
with history of falls and slow gait speed). In addition, among
subjects with history of previous falls, a faster gait speed (! 1 m/s)
was a significant protective factor for subsequent falls.
4. Discussion
4.1. Risk for falls among older people
This prospective cohort study evaluated the effectiveness of
adding gait speed to history of previous falls in predicting falls in
the subsequent year among male people aged 80 years and older in
Taiwan. Results of this study provided risk stratification for
octogenarians with the same history of falls in subsequent falls,
which highlighted the benefits of this combined approach to
improve identifying people with higher risk of falls. In this study,
26.1% of all study subjects fell at least once in the follow-up period.
Older age, history of falls in the previous year, and presence of pain
were all independent risk factors for falls in this study, which were
similar to previous studies. However, subjects with previous
history of fall plus slower gait speed (defined as < 0.5 m/s) were at
a very high risk of falls. On the contrary, subjects with their gait
speed ! 1 m/s were not at a higher risk of falls even if they had a
previous history of falls. These findings strengthened the AGS-BGS
guidelines in the fall prevention programs, and were a simple and
efficient approach to stratify risk of falls for people aged 80 years
and older.

Table 2
Independent risk factors for falls among men aged 80 years and older during 12-month follow-up.

Independent Variablesa

Unadjusted OR

95% CI

P value

Adjusted OR

95% CI

P value

History of falls in past 1 year
Age (years)
Presence of pain (yes)
Gait speed (m/s)
Gait speed with cut-off point < 1 m/s
Gait speed with cut-off point 0.8 m/s
Gait speed with cut-off point < 0.5 m/s

3.742
1.059
2.298
0.319
–
–
2.987

1.992–7.030
0.987
1.257–4.202
0.119–0.854
–
–
1.577–5.656

< 0.001
1.136
0.007
0.023
–
–
0.001

4.255
1.128
2.674
–
–
–
2.964

2.089–8.667
1.031–1.234
1.332–5.369
–
–
–

1.394–6.300

< 0.001
0.008
0.006
–
–
–
0.005

a
Covariates adjusting for age, polypharmacy, hearing impairment, urine incontinence, history of falling in past 1 year, EQ5D VAS scores, with symptoms of pain, IADL, and
depressive symptoms based on GDS-15, and gait speed (m/s).

Please cite this article in press as: Liang C-K, et al. Gait speed and risk assessment for falls among men aged 80 years and older: A
prospective cohort study in Taiwan. Eur Geriatr Med (2014), />

G Model

EURGER-511; No. of Pages 5
C.-K. Liang et al. / European Geriatric Medicine xxx (2014) xxx–xxx

4

Table 3
Synergistic effect of history of falls and slow gait speed in predicting subsequent falls.
Cut-off point of
slower gait speed < 1 m/s
Dependent variablesa
History

History
History
History

of
of
of
of

falls
falls
falls
falls

(–)
(–)
(+)
(+)

and
and
and
and

slow
slow
slow
slow

gait

gait
gait
gait

speed
speed
speed
speed

(–)
(+)
(–)
(+)

Cut-off point of
slower gait speed

Cut-off point of
slower gait speed < 0.5 m/s

0.8 m/s

Adjusted OR

95% CI

P value

Adjusted OR

95% CI

P value

Adjusted OR

95% CI

P value

Reference
–
–
3.052

–
–
–
1.054–8.840

–
–
–
0.040

Reference
–
4.290
4.044

–
–
1.379–13.346
1.504–10.869

–
–
0.012
0.006

Reference
3.308
4.423
10.920

–
1.280–8.549
1.857–10.535
3.423–34.839

–
0.014
0.001
< 0.001

a
Covariates adjusting for age, polypharmacy, hearing impairment, urine incontinence, EQ5D VAS scores, with symptoms of pain, IADL, and depressive symptoms based on
GDS-15.

4.2. Gait speed and falls

Slower gait speed has been recognized to be associated with
risk of falls for the elderly, which was a simple and effective test in
clinical practice [8,23,27,28]. The E´pidemiologie de l’Osteoporose
study, EPIDOS, surveyed 7575 community-dwelling French
women aged 75 years and older showing that the lowest quartile
of gait speed had a 1.4Â higher risk (95% CI 1.1–1.6) of femoral neck
fracture than the highest quartile during a mean 1.9 years followup [27]. Besides, Chu et al. enrolled 1517 elderly Hong Kong people
and disclosed that the faster gait speed was a significant protective
factor against falls [8]. Despite all the supporting evidences,
the cut-off of gait speed in predicting falls remained unclear,
especially for those who aged 80 years and older. Montero-Odasso
et al. reported that the gait speed < 0.7 m/s was significantly
predictive for falls in the coming two years (RR = 5.4, 95% CI: 2.0–
4.3) by a study sample with similar age to ours (102 participants
aged 75 years and older) [28]. Compared to the previous studies,
CGA was performed for all study subjects to identify potential
confounders in predicting subsequent falls, including visual and
hearing impairment, malnutrition, presence of pain, and comorbidities. In 2012, Taekema et al. found that the gait speed
of < 0.46 m/s in men was associated with a higher risk of mortality
during 12-year follow-up [14]. In this study, a gait speed
of < 0.5 m/s was a stronger predictive factor for falls than the
slower gait speed using the cut-off of 0.8 or 1 m/s.

physiological, functional and psychological conditions of old
women with a recent history of falls [32]. Results of this study
showed that older people with history of falls eventually were at
no higher risk for falls when their gait speed exceeded 1.0 m/s,
so they may not be the most prioritized population for fall
prevention programs even though they reported a previous
history of falls. In particular, those who had a previous history of

falls and the gait speed of < 0.5 m/s, a stronger intervention
program should be introduced to prevent subsequent falls
because they were at very high risk of falls.
4.4. Limitations
Despite all the research efforts went into this study, there were
still several limitations. First, the study sample was homogenous
in their demographic characteristics that they were all males, and
all veterans were living in the same retirement community.
However, we believe results of this study were still of great
implications to effectively conduct fall prevention programs in
the communities. Second, among all eligible study subjects, 27
refused the evaluation of EQ5D, and 61 of them refused being
measured for their BMI. These conditions may lessened the
statistic power, however, results of this study still clearly showed
the benefits of adding gait speed measurements to history of falls
in predicting falls. Third, a self-report bias of fall episodes may
exist in this study because older veterans may be reluctant to
report falls due to their self-esteem.

4.3. Combined approach of history of falls and gait speed
4.5. Conclusions
The ‘‘history of fall’’ was a strong risk factor for subsequent
falls traditionally, but physical conditions of older people with
the same history of falls may differ extensively. Rekeneire et al.
reported that fallers might eventually have certain subclinical
deficits that were overlooked by clinicians [29]. Moreover,
analysis of gait characteristics among older people with a
previous history of falls showed that their gait speed varied
extensively from 0.18 to 1.07 m/s [30], which clearly demonstrated the great heterogeneity of these people that were
classified as at the same risk for falls due to the history of

previous falls. Therefore, adding the dynamic gait speed
measurement to the static history of falls improved the
effectiveness in risk stratification for fall prevention programs.
In particular, history of falls is not a modifiable risk factor, but a
slower gait speed may be improved through certain intervention
programs. In this study, the history of fall remained to be a
strong independently risk factor for falls no matter slow gait
speed was defined as < 0.5 or 0.8 m/s. It has been reported
that adequate resistance training would improve the gait speed,
even among people aged 90 years and older [31], and the 6month multidimensional training program may sustain the
beneficial effects upon the increasing gait speed, as well as the

Among men aged 80 years and older living in the retirement
communities, the history of previous falls alone was not a single
best risk factor for subsequent falls. A gait speed ! 1 m/s was
protective against subsequent falls in spite of the presence of
history of previous falls. On the other hand, for subjects with the
history of previous falls and a slow gait speed (< 0.5 m/s), a
stronger intervention program should be introduced due to the
disproportionately high risk of fall in the following year.
Disclosure of interest
The authors declare that they have no conflicts of interest
concerning this article.
Acknowledgement
The study was supported by the Veteran Affairs Council, R.O.C
(Grant number: VAC101-C1 and VAC102-C1) and all authors
declare no conflicts of interest. The study group thanks all staff in
the Gangshan Veterans Home for their valuable assistance in
obtaining the information.

Please cite this article in press as: Liang C-K, et al. Gait speed and risk assessment for falls among men aged 80 years and older: A
prospective cohort study in Taiwan. Eur Geriatr Med (2014), />

G Model

EURGER-511; No. of Pages 5
C.-K. Liang et al. / European Geriatric Medicine xxx (2014) xxx–xxx

Ethical approval: The whole study has been approved by the Institutional Review Board of Kaohsiung Veterans General Hospital, Taiwan.

References
[1] Chen CY. Meeting the challenges of elder care in Taiwan’s aging society. J Clin
Gerontol Geriatr 2010;1:2–4.
[2] Sattin RW. Falls among older persons: a public health prospective. Ann Rev
Public Health 1992;13:489–508.
[3] Dellinger AM, Stevens JA. The injury problem among older adults: mortality,
morbidity and costs. J Safety Res 2006;37:519–22.
[4] Tinetti ME, Speechley M, Ginter SF. Risk factors for falls among elderly persons
living in the community. N Engl J Med 1988;319:1701–7.
[5] Tinetti ME, Baker DI, McAvay G, Claus EB, Garrett P, Gottschalk M, et al. A
multifactorial intervention to reduce the risk of falling among elderly people
living in the community. N Engl J Med 1994;331:821–7.
[6] Kenny RA, Rubenstein LZ, Tinetti ME, Brewer K, Cameron KA, Capezuti EA, et al.
Summary of the Updated American Geriatrics Society/British Geriatrics Society clinical practice guideline for prevention of falls in older persons. J Am
Geriatr Soc 2011;59:148–57.
[7] Bongue B, Dupre´ C, Beauchet O, Rossat A, Fantino B, Colvez A. A screening tool
with five risk factors was developed for fall-risk prediction in communitydwelling elderly. J Clin Epidemiol 2011;64:1152–60.
[8] Chu LW, Chi I, Chiu AY. Incidence and predictors of falls in the Chinese elderly.
Ann Acad Med Singapore 2005;34:60–72.
[9] Teno J, Kiel DP, Mor V. Multiple stumbles: a risk factor for falls in communitydwelling elderly: a prospective study. J Am Geriatr Soc 1990;38:1321–5.

[10] Edwards MH1, Jameson K, Denison H, Harvey NC, Sayer AA, Dennison EM, et al.
Clinical risk factors, bone density and fall history in the prediction of incident
fracture among men and women. Bone 2013;52:541–7.
[11] Abellan van Kan G, Rolland Y, Andrieu S, Bauer J, Beauchet O, Bonnefoy M, et al.
Gait speed at usual pace as a predictor of adverse outcomes in communitydwelling older people an International Academy on Nutrition and Aging
(IANA) Task Force. J Nutr Health Aging 2009;13:881–9.
[12] Cruz-Jentoft AJ, Baeyens JP, Bauer JM, Boirie Y, Cederholm T, Landi F, et al.
European Working Group on Sarcopenia in Older People Sarcopenia: European
consensus on definition and diagnosis: report of the European Working Group
on Sarcopenia in Older People. Age Ageing 2010;39:412–23.
[13] Studenski S, Perera S, Patel K, Rosano C, Faulkner K, Inzitari M, et al. Gait speed
and survival in older adults. JAMA 2011;305:50–8.
[14] Cesari M, Kritchevsky SB, Penninx BW, Nicklas BJ, Simonsick EM, Newman AB,
et al. Prognostic value of usual gait speed in well-functioning older people –
results from the Health Aging and Body Composition Study. J Am Geriatr Soc
2005;53:1675–80.

5

[15] Cesari M, Onder G, Zamboni V, Manini T, Shorr RI, Russo A, et al. Physical
function and self-rated health status as predictors of mortality: results from
longitudinal analysis in the ilSIRENTE study. BMC Geriatry 2008;8:34.
[16] Taekema DG, Gussekloo J, Westendorp RG, de Craen AJ, Maier AB. Predicting
survival in oldest old people. Am J Med 2012;125:1188e1–94e1.
[17] Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying
prognostic comorbidity in longitudinal studies: development and validation. J
Chronic Dis 1987;40:373–83.
[18] Liao YC, Yeh TL, Ko HC, et al. Geriatric depression scale (GDS) – a preliminary
report of reliability and validity for Chinese version. Changhua Med J
1995;1:11–7 [in Chinese].

[19] Kaiser MJ, Bauer JM, Ramsch C, Uter W, Guigoz Y, Cederholm T, et al. Validation
of the Mini Nutritional Assessment Short-Form (MNA-SF): a practical tool for
identification of nutritional status. J Nutr Health Aging 2009;13:782–8.
[20] Tombaugh TN, McIntyre NJ. The mini-mental state examination: a comprehensive review. J Am Geriatr Soc 1992;40:922–35.
[21] Lawton MP, Brody EM. Assessment of older people: self-maintaining and
instrumental activities of daily living. Gerontologist 1969;9:179–86.
[22] Rabin R, De Charro F. EQ-5D: a measure of health status from the EuroQol
Group. Ann Med 2001;33:337–43.
[23] Liu LK, Lee WJ, Liu CL, Chen LY, Lin MH, Peng LN, et al. Age-related skeletal
muscle mass loss and physical performance in Taiwan: implications to diagnostic strategy of sarcopenia in Asia. Geriatr Gerontol Int 2013;13:964–71.
[24] Lee WJ, Liu LK, Peng LN, Lin MH, Chen LK, Research ILAS Group. Comparisons of
sarcopenia defined by IWGS and EWGSOP criteria among older people: results
from the I-Lan longitudinal aging study. J Am Med Dir Assoc 2013;14:528e1–7.
[25] Biderman A, Cwikel J, Fried AV, Galisnky D. Depression and falls among
community dwelling elderly people: a search for common risk factors. J
Epidemiol Community Health 2002;56:631–6.
[26] A report of the Kellogg International Work Group on the prevention of falls by
the elderly. The prevention of falls in later life. Dan Med Bull 1987;34:1–24.
[27] Dargent-Molina P, Favier F, Grandjean H, et al. Fall-related factors and risk of
hip fracture: the EPIDOS prospective study. Lancet 1996;348:145–9.
[28] Montero-Odasso M, Schapira M, Soriano ER, Varela M, Kaplan R, Camera LA,
Mayorga LM. Gait velocity as a single predictor of adverse events in healthy
senior aged 75 years and older. J Gerontol Med Sci 2005;60:1304–9.
[29] De Rekeneire N, Visser M, Peila R, Nevitt MC, Cauley JA, Tylavsky FA, et al. Is a
fall just a fall: correlates of falling in healthy older persons. The Health, Aging
and Body Composition Study. J Am Geriatr Soc 2003;51:841–6.
[30] Barak Y, Wagenaar RC, Holt KG. Gait characteristics of elderly people with a
history of falls: a dynamic approach. Phys Ther 2006;86:1501–10.
[31] Idland G, Sylliaas H, Mengshoel AM, Pettersen R, Bergland A. Progressive
resistance training for community-dwelling women aged 90 or older: a

single-subject experimental design. Disabil Rehabil 2014;36(15):1240–8.
[32] Beyer N, Simonsen L, Bu¨low J, Lorenzen T, Jensen DV, Larsen L, et al. Old women
with a recent fall history show improved muscle strength and function
sustained for six months after finishing training. Aging Clin Exp Res
2007;19:300–9.

Please cite this article in press as: Liang C-K, et al. Gait speed and risk assessment for falls among men aged 80 years and older: A
prospective cohort study in Taiwan. Eur Geriatr Med (2014), />