Original research

Br J Sports Med: first published as 10.1136/bjsports-2021-105360 on 21 June 2022. Downloaded from http://bjsm.bmj.com/ on August 25, 2022 at University of Edinburgh. Protected by
Successful 10-­second one-­legged stance performance
predicts survival in middle-­aged and older individuals
Claudio Gil Araujo ‍ ‍,1 Christina Grüne de Souza e Silva ‍ ‍,1
Jari Antero Laukkanen ‍ ‍,2,3 Maria Fiatarone Singh ‍ ‍,4
Setor Kwadzo Kunutsor ‍ ‍,5,6 Jonathan Myers ‍ ‍,7 João Felipe Franca ‍ ‍,1
Claudia Lucia Castro ‍ ‍1

► Additional supplemental ABSTRACT balance tends to be reasonably preserved until the

material is published online Objectives Balance quickly diminishes after the mid-­ sixth decade of life, when comparatively, it starts to
only. To view, please visit the
journal online (http://d​ x.​doi.​ 50s increasing the risk for falls and other adverse health diminish quickly.15 16
org/1​ 0.​1136/​bjsports-​2021-​ outcomes. Our aim was to assess whether the ability to Nevertheless, balance assessment is not routinely
105360). complete a 10- s one-­legged stance (10-­second OLS) incorporated in the clinical examination of middle-­
1
is associated with all-­cause mortality and whether it aged and older individuals.17 This may be partly
Exercise Medicine Clinic-­ attributable to the poor standardisation of balance
adds relevant prognostic information beyond ordinary
CLINIMEX, Rio de Janeiro, Brazil
2
Institute of Clinical Medicine, demographic, anthropometric and clinical data. testing as well as to the relative paucity of data-­
Department of Medicine, Methods Anthropometric, clinical and vital status and relating balance results to clinical outcomes other
University of Eastern Finland, 10-­s OLS data were assessed in 1702 individuals (68% than falls, such as mortality, when compared with,
Kuopio, Finland men) aged 51–75 years between 2008 and 2020. Log-­ for example, aerobic fitness.16 18 19
3
Central Finland Health
Care District, Department of rank and Cox modelling were used to compare survival In this context, the availability of simple, inex-
Medicine, Jyväskylä, Finland curves and risk of death according to ability (YES) or pensive, reliable and safe balance assessment tools11
4
School of Health Sciences and inability (NO) to complete the 10-­s OLS test. that could help predict survival would potentially
Sydney Medical School, Faculty Results Overall, 20.4% of the individuals were be beneficial to health professionals evaluating
of Medicine and Health, The and treating older adults. Therefore, the aims of
classified as NO. During a median follow-­up of 7 years,
University of Sydney, Sydney,
New South Wales, Australia 7.2% died, with 4.6% (YES) and 17.5% (NO) on the our study were: (1) to assess whether the ability
5
National Institute for Health 10-­s OLS. Survival curves were worse for NO 10-­s to complete a 10-­s one-­legged stance (10-­s OLS)

copyright.
Research Bristol Biomedical OLS (log-­rank test=85.6; p<0.001). In an adjusted test was independently associated with all-­ cause
Research Centre, University model incorporating age, sex, body mass index and mortality in middle-­aged and older men and women
Hospitals Bristol and Weston
NHS Foundation Trust and the comorbidities, the HR of all-­cause mortality was higher and (2) whether the 10-­s OLS added relevant prog-
University of Bristol, Bristol, UK (1.84 (95% CI: 1.23 to 2.78) (p<0.001)) for NO nostic information beyond ordinary demographic,
6
Translational Health Sciences, individuals. Adding 10-­s OLS to a model containing anthropometric and clinical data. If the ability to
Bristol Medical School, established risk factors was associated with significantly perform this simple physical task were shown
University of Bristol, Southmead to be a good prognostic indicator for risk of all-­
improved mortality risk prediction as measured
Hospital, Bristol, UK
7
Cardiology Division, VA Palo by differences in −2 log likelihood and integrated cause mortality, it might be a useful complement to
Alto Health Care System and discrimination improvement. routine evaluations among middle-­aged and older
Stanford University, Palo Alto, Conclusions Within the limitations of uncontrolled subjects.
California, USA variables such as recent history of falls and physical
activity, the ability to successfully complete the 10-­s OLS
Correspondence to METHODS
is independently associated with all-­cause mortality and
Dr Claudio Gil Araujo, Clinimex
Medicina do Exercicio, Rio de adds relevant prognostic information beyond age, sex This was a prospective cohort study using data from
Janeiro 22031-­071, Brazil; and several other anthropometric and clinical variables. the CLINIMEX Exercise open cohort/evaluation
​cgaraujo@​iis.c​ om.​br There is potential benefit to including the 10-­s OLS as protocol (see online supplemental materials).11 20 21
part of routine physical examination in middle-­aged and Briefly, the CLINIMEX Exercise cohort study was
Accepted 19 April 2022
older adults. set up in 1994 to assess the relationships of various
Published Online First
21 June 2022 measures of physical fitness and other exercise-­
related variables, as well as conventional cardiovas-
cular risk factors with all-­cause and cause-­specific
INTRODUCTION mortality outcomes. The sample size of 1593
Ageing is associated with a progressive decline in participants was calculated based on the following
physical fitness1–3 and reductions or impairments in parameter specifications: (1) level of significance,
components of aerobic4 5 and non-­aerobic fitness, two-­ sided test at α=0.05; (2) power (1−β) of
including muscle strength/power, flexibility, balance 80%; (3) 7% of study participants dying during
© Author(s) (or their
employer(s)) 2022. No and body composition.6–11 It is also well-­established follow-­up; (4) an SD of 0.5 for the exposure (given
commercial re-­use. See rights that the combination of sarcopenic obesity and loss that the binary exposure follows a Bernoulli distri-
and permissions. Published of flexibility and balance are detrimental for overall bution with the probability of a subject achieving
by BMJ. health, placing older adults with frailty more prone success, p, assumed to be equal to 0.5, the SD was
To cite: Araujo CG, de Souza to falls and other serious adverse medical sequelae.12 calculated from the formula: (p*(1−p))0.5 and (5)
e Silva CG, Laukkanen JA, Indeed, falls are the second leading cause of unin- effect size: the minimum HR considered to be clin-
et al. Br J Sports Med tentional injury-­based deaths worldwide.13 Unlike ically important, in this case, 1.7. The current anal-
2022;56:975–980. aerobic fitness,2 14 muscle strength9 and flexibility,8 ysis included 1702 participants aged 51–75 years at

Araujo CG, et al. Br J Sports Med 2022;56:975–980. doi:10.1136/bjsports-2021-105360    1 of 7

 Original research

Br J Sports Med: first published as 10.1136/bjsports-2021-105360 on 21 June 2022. Downloaded from http://bjsm.bmj.com/ on August 25, 2022 at University of Edinburgh. Protected by
their first evaluation conducted between 10 February 2009 and
10 December 2020, who voluntarily sought the clinic for evalu-
ation to assess aerobic and non-­aerobic physical fitness and/or to
obtain exercise counselling.
Censoring of vital status and mortality was updated to
mid-­ December 2020 from the regional official registry data.
All participants read and signed an informed consent and
formally authorised the use of their deidentified data for scien-
tific purposes. The study protocol was included in a National
Research Registry and formally reviewed and approved by an
external Research Ethics Committee (Plataforma Brasil—CAAE:
40122320.8.0000.9433). Cohort data are maintained in an
institutional database.

Demographics, anthropometric and clinical variables

Sex, age, date of evaluation, date of death (if it occurred),
censored date and follow-­up time were available for all partici-
pants. Anthropometric measurements were obtained with partic-
ipants barefoot and using minimal clothing and included height,
weight, sum of six skinfolds—triciptal, subscapular, suprailiac,
abdominal, thigh and medial calf - and waist girth measured at
the umbilical level, as well as two calculated values—body mass
index (BMI) and waist-­height ratio. Clinical data were obtained
by medical history, considering the presence or absence of known
relevant diseases and/or use of regular medications. Obesity was
defined as a BMI ≥30 kg/m2. Less than 1% of anthropometric
or clinical data were missing. All participants were fully ambu-
latory and those presenting with unstable gait or having signs of
any known acute vestibular or otoneurological disturbance were
excluded.

copyright.
One-legged stance balance assessment
As part of the evaluation, participants were asked to stand on
a flat platform. Static balance was assessed as the ability to
complete 10 s in OLS, either left or right foot, under close face-­
to-­face supervision of a physician and/or a nurse assistant as a
precaution to prevent falls or injuries. To minimise the influence
of muscle strength and flexibility and to improve standardisa-
tion, barefoot participants were instructed to place the dorsal
part of the non-­support foot on the back of the opposite lower
leg, as naturally as possible (figure 1). Additionally, participants
were asked to keep their elbows extended, the arms naturally
placed close to their body and instructed to fix their gaze on an
eye-­level point at a 2-­m distance.22 Once the participant assumed
the correct position, a count of 10 s was started and up to three Figure 1 Body position to the 10-­s one-­legged stance test.
attempts were allowed. A very simple criterion was applied—
ability to complete 10-­s OLS on either foot, keeping the correct count less than 1 and no more than 20% of cell counts less than
initial position and without any other support—and participants 5, depending on the variable. Spearman rank correlation coeffi-
were accordingly classified as ability (YES) or inability (NO) to cients were used to calculate associations. Kaplan-­Meier curves
complete the 10-­s OLS test. were constructed and log-­rank tests were used to analyse survival
times for the YES and NO 10-­s OLS groups.
Statistical analysis While the CLINIMEX Exercise cohort comprises men
For descriptive statistics, quantitative variables were described and women from 6 to 99 years of age, only participants aged
using mean±SD or median and IQR, depending on the nature between 51 and 75 years were included in this mortality study,
of distribution, and categorical variables were summarised as survival curves (see online supplemental materials) calculated
using frequencies and percentages. Sample size calculations at each 5 years of age interval starting at 41–45 years indicated
employed the Stata command “stpower cox” which imple- that the combination of relevant numbers of deaths and failures
ments the methods of Hsieh and Lavori23 and Schoenfeld.24 The to complete the 10-­s OLS test could be identified only using the
Cochran–Armitage test was used to test for temporal trends in 51–75 years of age range. The relationship between 10-­s OLS
deaths across the follow-­up period. Results for YES and NO 10-­s results and all-­cause mortality was modelled by Cox univariate
OLS were compared by two-­tailed Student’s t-­tests or χ2 test, and multivariable analyses, after confirmation of no departure
after checking, respectively, for the normality of distribution and from the proportionality of hazards assumptions using Schoen-
homogeneity of variance and for the inexistence of expected cell feld residuals.25 The proportionality test of each covariate as well

2 of 7 Araujo CG, et al. Br J Sports Med 2022;56:975–980. doi:10.1136/bjsports-2021-105360

 Original research

Br J Sports Med: first published as 10.1136/bjsports-2021-105360 on 21 June 2022. Downloaded from http://bjsm.bmj.com/ on August 25, 2022 at University of Edinburgh. Protected by
as a global test was done. The test was not statistically significant
for each of the covariates, and the global test was also not statis-
tically significant. Adjustments were made for age, sex, BMI and
clinical variables (as previously described), using the 10-­s OLS
YES group as the reference. For each participant, follow-­up time
was obtained using the number of days between the evaluation
and death or censoring dates. None of the participants were lost
to follow-­up.
To assess whether adding information on 10-­s OLS to conven-
tional risk factors was associated with improvement in the
prediction of all-­ cause mortality, three statistical approaches
were employed. First, the improvement in risk discrimination
resulting from adding information on 10-­ s OLS to a model
containing established risk factors (age, sex, BMI, medical history
of coronary artery disease, hypertension, diabetes and dyslipi-
daemia) was quantified using Harrell’s C-­index.26 Comparison
of the C-­index for models including and not including informa-
tion on the 10-­s OLS was performed according to the method-
ology of DeLong.27 The 95% CIs for C-­indices and their changes
were derived from jackknife SE. The C-­index is appropriate for
time-­to-­event data and provides the probability that the model Figure 2 YES= ability or NO= inability to complete the test to 10-­s
correctly predicts the order of failure of randomly selected pairs one-­legged stance test according to age groups. This figure includes
of individuals. A C-­index of 1.0 indicates perfect prediction of information from individuals of a wider age range than the one included
the order of failure (in this case, mortality), whereas a C-­index for analysis in this study as mentioned previously.
of 0.5 is achieved purely by chance. Second, the continuous
net reclassification improvement (NRI) was calculated,28 which
determines whether risk increases cases applying a new model all-­cause mortality events, we had 92% power to detect a clin-
compared with an established or reference model. Additionally, ically important HR of 1.84. The proportion of deaths in the
the integrated discrimination improvement (IDI) was calculated, NO group was higher than that in the YES group (17.5% vs
which integrates the NRI over all possible cut-­offs.29 In addition 4.6%; p<0.001), reflecting an absolute difference of 12.9%, but
to Harrel’s C-­index which can be insensitive in detecting differ- the distribution of the major underlying causes of death did not

copyright.
ences because it is based on ranks rather than on continuous differ significantly between the YES and NO groups (p=0.45).
data and not being able to assess calibration,30 31 we tested for A comparison of key variables for all participants and separately
differences in the −2 log likelihood of prediction models with for 10-­s OLS YES and NO groups is presented in table 1. While sex
and without inclusion of 10-­s OLS. The −2 log likelihood test distribution did not different significantly between the YES and NO
has been recommended as a more sensitive risk discrimination groups, (p=0.76), several other variables including age, BMI and
method.30 31 Statistical significance level was set at 5%, and 95% waist-­height ratio differed between the two groups (p<0.001). In
CIs were calculated for all results. Calculations were performed general, NO participants had an unhealthier profile with a higher
and figures prepared using either Prism (V.8.4.3; GraphPad, percentage of participants having coronary artery disease, hyper-
USA) or STATA (V.16; USA) statistical packages. tension, dyslipidaemia and obesity. The most striking difference
was for diabetes mellitus, which was three times more common in
the NO group (37.9%) as compared with the YES group (12.6%)
Patient and public involvement (p<0.001).
This research was done without patient or public involvement in Correlation coefficients between a NO 10-­s OLS response and age
the study design, data analysis, writing or editing. and several anthropometric variables of potential interest are shown
in figure 3. With the exception of height, inability to complete the
RESULTS 10-­s OLS was significantly associated with all the other variables
The mean±SD age of the participants was 61.7±6.8 years and (p<0.001). The two highest correlation coefficients between NO
68% were men. No adverse medical events or accidents occurred and selected continuous variables were 0.40 for age and 0.26 for
during the 10-­s OLS testing. A total of 348 (20.4%) participants waist-­height ratio.
failed to pass the test and were classified as NO. The inability
to complete the test, that is, 10-­s OLS with either the right or 10-s OLS and all-cause mortality association and risk
left foot, increasing with aging, practically doubling at each prediction
subsequent 5-­year age-­group intervals beginning at age group Kaplan-­Meier survival curves were significantly different for YES
of 51–55 years. The proportion of NO responders was 4.7% and NO responders (p<0.001) (figure 4). Cox proportional hazard
among those 51–55 years, 8.1% at 56–60 years, 17.8% at 61–65 analysis indicated that inability to complete the 10-­s OLS was asso-
years and 36.8% at 66–70 years. In the age group of 71–75 ciated with a significantly higher risk for all-­cause mortality. An age-­
years, the majority of the participants (53.6%) were unable to adjusted and a multivariable-­adjusted—age, sex, BMI and clinical
successfully complete the 10-­s OLS (figure 2). During a median comorbidities (including history of coronary artery disease, hyper-
(IQR) follow-­up time of 7 (4.16–9.41) years, 123 participants tension, diabetes, obesity and dyslipidaemia) showed HRs (95% CI)
(7.2%) died, mostly due to cancer (32%), cardiovascular causes of 2.18 (1.48 to 3.22; p<0.001) and 1.84 (1.23 to 2.78; p=0.003)),
(30%), diseases of the respiratory system (9%) and COVID-­19 respectively (table 2). A directed acyclic graph showing a mini-
complications (7%) with no clear temporal trends in the deaths mally sufficient set of confounders for adjustment is also presented
(p=0.77). Given a sample of 1702 individuals including 123 (figure 5).

Araujo CG, et al. Br J Sports Med 2022;56:975–980. doi:10.1136/bjsports-2021-105360 3 of 7

 Original research

Br J Sports Med: first published as 10.1136/bjsports-2021-105360 on 21 June 2022. Downloaded from http://bjsm.bmj.com/ on August 25, 2022 at University of Edinburgh. Protected by
Table 1 Demographic and clinical characteristics of men and women aged 51–75 years according to the ability to complete 10-­s one-­legged
stance test
10-­s one-­leg stance test

Variable* Total (N=1702) Yes (n=1354) No (n=348) P value†

Age (years) 61.7±6.8 60.3±6.2 67.2±6.0 <0.001
Sex (male) 67.9 68.1 67.2 0.761
Weight (kg) 79.9±16.0 79.0±15.5 83.6±17.3 <0.001
Height (cm) 169.9±9.2 171.1±9.1 169.1±9.6 0.062
Body mass index (kg/m2) 27.6±4.5 27.2±4.2 29.2±5.2 <0.001
Waist-h­ eight ratio 0.57±0.07 0.56±0.07 0.61±0.08 <0.001
Sum of skinfolds (mm)‡ 117.7±1.0 114.6±1.1 129.7±2.4 <0.001
Comorbidities
 Hypertension (%) 47.9 43.5 65.3 <0.001
 Dyslipidaemia (%) 54.8 52.7 63.0 0.001
 Diabetes mellitus (%) 17.7 12.6 37.9 <0.001
 Obesity (%) 26.2 22.6 40.2 <0.001
 Coronary artery disease (%) 32.1 30.0 40.5 <0.001
 Myocardial infarction (%) 16.1 15.4 18.8 0.124
 CABG (%) 8.7 7.5 13.3 0.001
 PCI (%) 21.7 20.7 25.4 0.057
Death (%) 7.2 4.6 17.5 <0.001
Follow-­up time (days) 2538 (1518–3434) 2628 (1594–3491) 2123 (1146–3156) <0.001
*Mean±SD or %.
†Student’s t-­test or χ2 test.
‡Sum of six skinfolds—triciptal, subscapular, suprailiac, abdominal, anterior thigh and medial calf.
CABG, coronary artery bypass graft; PCI, percutaneous coronary intervention.

copyright.
An all-­cause mortality model containing established risk factors respectively, suggesting additive value of the inability to complete
yielded a C-­index of 0.7990 (0.7563, 0.8417). After addition of the the 10-­s OLS test.
10-­s OLS binary results, the C-­index was 0.8090 (0.7678, 0.8503),
an increase of 0.0100 (–0.0005, 0.0205; p=0.06). The −2 log like- DISCUSSION
lihood model showed significant improvement after addition of the Each year an estimated 684 000 individuals die from falls globally,
10-­s OLS binary results to the model (p for comparison=0.002). of which over 80% are in low/middle-­income countries.13 While
The continuous NRI and IDI were 17.50% (95% CI −1.46 to it is known that good levels of balance are relevant for many daily
36.45; p=0.07) and 0.0143 (95% CI 0.0019 to 0.0267; p=0.024), life activities,32 there is considerable evidence that loss of balance
is also detrimental for health and that some exercise interventions
may improve balance.12 32–37 However, it is currently uncertain
if the results of repeated 10-­s OLS tests would be amenable to

Figure 3 Spearman correlation coefficients between the inability

to complete the 10-­s one-­legged stance test and demographic and Figure 4 Kaplan-­Meier survival curves of participants aged 51–75
anthropometric variables. Values of r>0.048 (positive or negative) were years old divided by ability (YES) and inability (NO) to complete the 10-­s
significant at 5% of probability. one-­legged stance test.

4 of 7 Araujo CG, et al. Br J Sports Med 2022;56:975–980. doi:10.1136/bjsports-2021-105360

 Original research

Br J Sports Med: first published as 10.1136/bjsports-2021-105360 on 21 June 2022. Downloaded from http://bjsm.bmj.com/ on August 25, 2022 at University of Edinburgh. Protected by
Prognostic information
Table 2 Associations of ability to complete 10-­s one-­legged stance
The ability to complete the 10-­s OLS test starts to progressively
balance test (10-­s OLS) with all-­cause mortality in 1702 men and
diminish with ageing, approximately halving at each subsequent
women aged 51–75 years old (123 deaths, 7.2%; median follow-­up
5-­year age group interval. Stated differently, the participants in the
time=7 years)
oldest age group—71–75 years—were more than 11 times more
Events/ All-­cause mortality likely to be a NO responder as compared with those just 20 years
10-­s OLS participants HR P value*
younger and belonging to the youngest age group in the study—
Model 1—unadjusted 51–55 years. The ability to complete the 10-­s OLS tended to show
 YES 62/1354 1 (Reference) both a ceiling and a floor in terms of an age profile, with very rare
 NO 61/348 4.58 (3.21 to 6.53) <0.001 (<1%) younger participants (<45 years of age) failing and relatively
Model 2—adjusted by age few participants older than 80 years able to complete the test (see
 YES 62/1354 1 (Reference) online supplemental material).
 NO 61/348 2.18 (1.48 to 3.22) <0.001 Univariate analysis indicated that a NO 10-­s OLS response was
Model 3—adjusted by age, significantly and directly associated with age, with a high waist-­
sex, BMI and comorbidities† height ratio and the prevalence of diabetes mellitus. Our results
 YES 62/1354 1 (Reference) are concordant with those of Neri et al40 who found that adiposity
 NO 61/348 1.84 (1.22 to 2.77) 0.003 measures, in particular waist circumference, were associated with
*Cox proportional hazards modelling. postural instability and higher risk of falls in older adults. In addi-
†Model 3: adjusted by age, sex, BMI and presence of coronary artery disease, tion, the higher percentage of participants with diabetes mellitus in
systemic arterial hypertension, dyslipidaemia and diabetes mellitus. the NO 10-­s OLS group suggests that some of these participants
BMI, body mass index. have subclinical central or autonomic nervous system dysfunction,
as has been recently reported.41
Our data show that middle-­aged and older participants unable to
intervention, that is, exercise or balance training, and if changes in complete the 10-­s OLS had lower survival over a median of 7 years
10-­s OLS over time would influence mortality risk.38 compared with those able to complete the test, with an 84% higher
In our 13 years of clinical experience routinely using the 10-­s risk of all-­cause mortality, even when other potentially confounding
OLS static balance test in adults with a wide age range and diverse variables such as age, sex, BMI and clinical comorbidities or risk
clinical conditions,39 the test has been remarkably safe, well-­ factors, including presence of coronary artery disease, hyperten-
received by the participants, and importantly, simple to incorpo- sion, obesity, dyslipidaemia and diabetes mellitus, were taken into
account. The utility of the 10-­s OLS test for mortality risk assessment
rate in our routine practice as it requires less than 1 or 2 min to be
is further corroborated by the fact that it provided an improvement

copyright.
applied.
in mortality risk discrimination using measures including IDI and
difference in −2 log likelihood.

Comparison of current findings with the literature

A study published in 200715 proposed normative values for OLS
timing based on results obtained in 549 men/women divided into
six age groups ranging from 18 to 80+ years. Similar to our study,
they found that OLS performance was strongly and negatively influ-
enced by age but unaffected by sex.15 There is limited information
in the literature relating balance to all-­cause mortality. In a recent
Japanese study with 1085 elderly participants (65–89 years),22 it
was observed that OLS timing was strongly associated with all-­cause
mortality, with an adjusted-­relative risk value similar to the current
study using similar covariates (1.91 (95% CI 1.39 to 2.63)). In a
recent cohort study, Cao et al42 evaluated static balance in 5816 men
and women older than 40 years who were followed for a median
of 12.5 years and observed that those with a balance disorder had
a 44% higher risk of all-­cause mortality when compared with those
having normal results in the four conditions assessed in the modified
Figure 5 Directed acyclic graph (DAG) of association between Romberg test of standing balance on firm and compliant support
demographic and anthropometric variables, presence of comorbidities, surfaces.
the ability to complete 10-­s one-­legged stance (10-­s OLS) and all-­cause It should be pointed out that OLS has been used to assess balance
mortality. The green box is the exposure variable (10-­s OLS). The blue for more than 50 years; Fregly et al16 were the first to report norma-
box ‘I’ is the outcome variable (all-­cause mortality). The grey boxes tive standards for OLS based on a healthy sample of military men
are potential cofounding variables. Other potential but unmeasured and women. It is notable that in most studies OLS results were
confounders were omitted from the figure for simplicity. Gender, 10-­s expressed as time in seconds, often limited to 30 or 60 s, reflecting
OLS, comorbidities, clinical status (healthy or unhealthy) and vital status the duration that participants were able to maintain the OLS posi-
were used as dichotomous variables. This DAG shows four confounding tion.15 16 43–45 While this is possible for young adults, it becomes
paths: the path via age, sex, body mass index (BMI) and clinical status progressively more difficult with ageing. Indeed, our data indicate
(healthy vs unhealthy). Thus, the minimally sufficient adjustment to get that most of participants aged >70 years were unable to complete
the total effect of the exposure on the outcome is to adjust for age, sex, 10-­s OLS. Moreover, while it seems that reliability is moderate to
BMI and clinical status. good for timing OLS studies,46 it is possible that in a clinical setting

Araujo CG, et al. Br J Sports Med 2022;56:975–980. doi:10.1136/bjsports-2021-105360 5 of 7

 Original research

Br J Sports Med: first published as 10.1136/bjsports-2021-105360 on 21 June 2022. Downloaded from http://bjsm.bmj.com/ on August 25, 2022 at University of Edinburgh. Protected by
with older participants timing measurements tend to be less repro-
ducible due to high intraparticipant and interobserver variation, KEY MESSAGES
potentially limiting the validity of the OLS results. Additionally,
WHAT ARE THE FINDINGS?
there are distinct ways in which OLS has been assessed, with vari-
ations in arm/hand positions, whether arm movements for stabili- ⇒ Ability to complete a 10s one-­legged stance (10-­s OLS) in
sation are allowed and position of the opposite leg and foot, with middle-­aged and older participants added relevant prognostic
some of the studies allowing a swing leg that incorporates a muscle information beyond ordinary clinical data.
strength component.18 38 47–49 HOW MIGHT IT IMPACT ON CLINICAL PRACTICE IN THE
Therefore, our results are not only confirmatory of the studies FUTURE?
mentioned above, but they extend these observations and make
⇒ 10-­s OLS provides rapid and objective feedback for the
them practical for routine clinical use. Indeed, it is simpler and likely
more reproducible to have a clear time reference such as 10 s as used
patient and health professionals regarding static balance.
⇒ The routine application of a simple and safe static balance
in the current study, as compared with recording the time in which
the subject is able to remain in OLS.
test—10-­s OLS—adds useful information regarding mortality
risk in middle-­aged and older men and women.

Strengths
There are several strengths to our study. The 10-­s OLS testing was
carried out in well-­controlled situations and under direct health CONCLUSION
professional surveillance. The 10-­s OLS test was easy to explain Our study indicates that the inability to complete a 10-­s OLS in
to the participant, to apply and to obtain the binary result used in middle-­aged and older participants is related to a higher risk of all-­
the study (YES or NO responder). Median follow-­up time and the cause mortality and, consequently, to a shorter life expectancy.
percentage of deaths in the sample provided appropriate statistical
power for analysis and to address the aim of the study. In addition, Correction notice This article has been corrected since it published Online First.
there were data available from several anthropometric and clin- The author’s name, Setor Kwadzo Kunutsor, has been amended and affiliations have
been updated.
ical variables that were used for clinically relevant multivariable
modelling. Twitter Claudio Gil Araujo @cgsaraujo
Acknowledgements The authors thank all participants who voluntarily permitted
us to anonymously use their data and to the Secretary of Health of Rio de Janeiro
Limitations state for providing the vital data information to the participants of CLINIMEX
Some limitations are notable. First, this CLINIMEX Exercise cohort Exercise cohort.
is primarily composed of participants of white race and belonging

copyright.
Contributors CGA, CGSS, CLC and JFF were involved in the planning of the study
to higher socioeconomic strata in Brazil. Any extrapolation of these and collecting data. Statistical analysis: CGA, CGSS, SK, JAL. Interpreting data: CGA,
findings to populations distinct from this profile should be inter- CGSS, MFS, JM, SK, JAL. Manuscript writing and revising: all authors. CGA acts as
preted with caution. It is also possible that a more sophisticated the guarantor of the study.
measure of OLS balance, such as centre of pressure displacement Funding CGA was partially sponsored by research grants from national and local
in a given period of time, would provide better discrimination and governmental agencies. Partial financial support was provided by CNPq e FAPERJ
research agencies.
improve the value of this assessment in terms of survival. However,
this would make testing much more difficult to incorporate as a Competing interests None declared.
simple clinical routine practice. Second, use of HRs may be unin- Patient and public involvement Patients and/or the public were not involved in
formative because they make direct comparisons between risk the design, conduct, reporting, or dissemination plans of this research.
factors difficult to interpret. They also implies a constant relative Patient consent for publication Not required.
hazard throughout the follow-­up, which is usually not the case.50 Ethics approval This study involves human participants and was approved by
In addition, they have a built-­in selection as a result of conditioning the Ethics Committee CONEP Brazilian Government (reference no: 4.459.555).
on those who have survived. Third, several potential confounder Participants gave informed consent to participate in the study before taking part.
variables were not available for the participants, including recent Provenance and peer review Not commissioned; externally peer reviewed.
history of falls, pattern of physical activity or exercise and sports Data availability statement Data are available on reasonable request.
practice, diet, smoking and the use of medications that may interfere Deidentified data are available on reasonable request.
with balance. Fourth, since we did not have repeat measurements of Supplemental material This content has been supplied by the author(s).
the exposure, we were unable to address time-­varying confounding It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not
as well as time-­varying confounding affected by prior exposure as have been peer-­reviewed. Any opinions or recommendations discussed are
potential confounders and to provide us the chance to correct for solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all
liability and responsibility arising from any reliance placed on the content.
potential regression dilution bias and, finally, we have not used the Where the content includes any translated material, BMJ does not warrant the
K-­fold cross-­validation’s approach, due to the relatively limited accuracy and reliability of the translations (including but not limited to local
sample size of our study. Future studies should explore whether 10-­s regulations, clinical guidelines, terminology, drug names and drug dosages), and
OLS results add prognostic information when data are available for is not responsible for any error and/or omissions arising from translation and
other components of physical fitness. adaptation or otherwise.
Finally, investigation of the biological mechanisms that may ORCID iDs
explain the observed associations between poor OLS balance Claudio Gil Araujo http://orcid.org/0000-0001-6679-6695
and all-­cause mortality is required. It is also of interest to investi- Christina Grüne de Souza e Silva http://orcid.org/0000-0002-0758-5174
gate whether more detailed or sophisticated assessments of static Jari Antero Laukkanen http://orcid.org/0000-0002-3738-1586
Maria Fiatarone Singh http://orcid.org/0000-0002-1897-8707
balance, such as including a measurement of the centre of pressure
Setor Kwadzo Kunutsor http://orcid.org/0000-0002-2625-0273
displacement, number of trials required, different arm or foot posi- Jonathan Myers http://orcid.org/0000-0003-2592-136X
tions and/or using closed eyes during the OLS, could contribute to João Felipe Franca http://orcid.org/0000-0001-6681-4152
even more powerful survival analyses. Claudia Lucia Castro http://orcid.org/0000-0002-6833-0407

6 of 7 Araujo CG, et al. Br J Sports Med 2022;56:975–980. doi:10.1136/bjsports-2021-105360

 Original research

Br J Sports Med: first published as 10.1136/bjsports-2021-105360 on 21 June 2022. Downloaded from http://bjsm.bmj.com/ on August 25, 2022 at University of Edinburgh. Protected by
REFERENCES 28 Pencina MJ, D’Agostino RB, Steyerberg EW. Extensions of net reclassification
1 Bull FC, Al-­Ansari SS, Biddle S, et al. World Health organization 2020 guidelines on improvement calculations to measure usefulness of new biomarkers. Stat Med
physical activity and sedentary behaviour. Br J Sports Med 2020;54:1451–62. 2011;30:11–21.
2 Peterman JE, Arena R, Myers J, et al. Development of global reference standards for 29 Pencina MJ, D’Agostino RB, D’Agostino RB, et al. Evaluating the added predictive
directly measured cardiorespiratory fitness: a report from the fitness registry and ability of a new marker: from area under the ROC curve to reclassification and
importance of exercise national database (Friend). Mayo Clin Proc 2020;95:255–64. beyond. Stat Med 2008;27:157–72.
3 Piercy KL, Troiano RP, Ballard RM, et al. The physical activity guidelines for Americans. 30 Cook NR. Use and misuse of the receiver operating characteristic curve in risk
JAMA 2018;320:2020–8. prediction. Circulation 2007;115:928–35.
4 Myers J, Prakash M, Froelicher V, et al. Exercise capacity and mortality among men 31 Harrell FEJ. Regression modeling strategies. New York: Springer, 2001.
referred for exercise testing. N Engl J Med 2002;346:793–801. 32 Keating CJ, Cabrera-­Linares JC, Párraga-­Montilla JA, et al. Influence of Resistance
5 Laukkanen JA, Lakka TA, Rauramaa R, et al. Cardiovascular fitness as a predictor of Training on Gait & Balance Parameters in Older Adults: A Systematic Review. Int J
mortality in men. Arch Intern Med 2001;161:825–31. Environ Res Public Health 2021;18. doi:10.3390/ijerph18041759. [Epub ahead of
6 Araújo CGS, Castro CLB, Franca JFC, et al. Sitting-­rising test: sex- and age-­reference print: 11 02 2021].
scores derived from 6141 adults. Eur J Prev Cardiol 2020;27:888–90. 33 Gonzalez AM, Mangine GT, Fragala MS, et al. Resistance training improves single leg
7 Laukkanen JA, Voutilainen A, Kurl S, et al. Handgrip strength is inversely associated stance performance in older adults. Aging Clin Exp Res 2014;26:89–92.
with fatal cardiovascular and all-­cause mortality events. Ann Med 2020;52:109–19. 34 Verhagen E, Bobbert M, Inklaar M, et al. The effect of a balance training
8 Medeiros HBdeO, de Araújo DSMS, de Araújo CGS. Age-­related mobility loss is joint-­ programme on centre of pressure excursion in one-­leg stance. Clin Biomech
specific: an analysis from 6,000 Flexitest results. Age 2013;35:2399–407. 2005;20:1094–100.
9 Vianna LC, Oliveira RB, Araújo CGS. Age-­related decline in handgrip strength differs 35 Muir SW, Berg K, Chesworth B, et al. Quantifying the magnitude of risk for balance
according to gender. J Strength Cond Res 2007;21:1310–4. impairment on falls in community-­dwelling older adults: a systematic review and
10 Araújo CGSde. Flexibility assessment: normative values for flexitest from 5 to 91 years meta-a­ nalysis. J Clin Epidemiol 2010;63:389–406.
of age. Arq Bras Cardiol 2008;90:257–63. 36 Hu MH, Woollacott MH. Multisensory training of standing balance in older adults: I.
11 Brito LBBde, Ricardo DR, Araújo DSMSde, et al. Ability to sit and rise from the floor as postural stability and one-­leg stance balance. J Gerontol 1994;49:M52–61.
a predictor of all-­cause mortality. Eur J Prev Cardiol 2014;21:892–8. 37 Mortazavi H. Balance training as a promising intervention for reducing the risk of fall
12 Izquierdo M, Merchant RA, Morley JE, et al. International exercise recommendations in elderly patients with chemotherapy-­induced peripheral neuropathy. J Cancer Res
in older adults (ICFSR): expert consensus guidelines. J Nutr Health Aging Ther 2019;15:1192–3.
2021;25:824–53. 38 Jaque-­Gallardo C, Véliz-­Campillay P, Cancino-­López J. [Effect of a high-­speed
13 WHO. Falls, 2021. Available: https://www.who.int/news-room/fact-sheets/detail/falls bodyweight resistance training on timed up and go and one leg stance in older
[Accessed 09 Nov 2021]. women]. Rev Med Chil 2019;147:1136–43.
14 Laukkanen JA, Zaccardi F, Khan H, et al. Long-­term change in cardiorespiratory 39 Araújo CG, SouzaeSilva CG, Castro CL. Mortality during the COVID-­19 pandemic:
fitness and all-­cause mortality: a population-­based follow-­up study. Mayo Clin Proc findings from the CLINIMEX exercise cohort in the year of 2020. medRxiv 2021.
2016;91:1183–8. 40 Neri SGR, Gadelha AB, de David AC, et al. The association between body adiposity
15 Springer BA, Marin R, Cyhan T, et al. Normative values for the unipedal stance test measures, postural balance, fear of falling, and fall risk in older community-­dwelling
with eyes open and closed. J Geriatr Phys Ther 2007;30:8–15. women. J Geriatr Phys Ther 2019;42:E94–100.
16 Fregly AR, Smith MJ, Graybiel A. Revised normative standards of performance of men 41 Yokoyama H, Shiraiwa T, Takahara M, et al. Applications of physical performance
on a quantitative ataxia test battery. Acta Otolaryngol 1973;75:10–16. measures to routine diabetes care for frailty prevention concept: fundamental data
17 Leroy V, Chen Y, Demnitz N, et al. Is fall risk systematically evaluated in memory with grip strength, gait speed, timed chair stand speed, standing balance, and knee

copyright.
clinics? a national survey of practice in France. J Alzheimers Dis 2021;81:1483–91. extension strength. BMJ Open Diabetes Res Care 2020;8:e001562.
18 Jonsson E, Seiger A, Hirschfeld H. One-­leg stance in healthy young and elderly adults: 42 Cao C, Cade WT, Li S, et al. Association of balance function with all-­cause and
a measure of postural steadiness? Clin Biomech 2004;19:688–94. cause-­specific mortality among US adults. JAMA Otolaryngol Head Neck Surg
19 Roeing KL, Hsieh KL, Sosnoff JJ. A systematic review of balance and fall risk 2021;147:460–8.
assessments with mobile phone technology. Arch Gerontol Geriatr 2017;73:222–6. 43 Hurvitz EA, Richardson JK, Werner RA, et al. Unipedal stance testing as an indicator of
20 Duarte CV, Myers J, de Araújo CGS. Exercise heart rate gradient: a novel index to fall risk among older outpatients. Arch Phys Med Rehabil 2000;81:587–91.
predict all-­cause mortality. Eur J Prev Cardiol 2015;22:629–35. 44 Ikezoe T, Tsuboyama T, Tabara Y, et al. Weak hip flexor strength predicts progression
21 Ramos PS, Araújo CGS. Cardiorespiratory optimal point during exercise testing as a of functional capacity decline due to locomotor system dysfunction in community-­
predictor of all-­cause mortality. Rev Port Cardiol 2017;36:261–9. dwelling older adults: a longitudinal cohort study. Arch Gerontol Geriatr
22 Nofuji Y, Shinkai S, Taniguchi Y, et al. Associations of walking speed, grip strength, and 2021;97:104499.
standing balance with total and cause-­specific mortality in a general population of 45 da Silva RA, Bilodeau M, Parreira RB, et al. Age-­related differences in time-­limit
Japanese elders. J Am Med Dir Assoc 2016;17:184.e1–184.e7. performance and force platform-­based balance measures during one-­leg stance. J
23 Hsieh FY, Lavori PW. Sample-­size calculations for the COX proportional hazards Electromyogr Kinesiol 2013;23:634–9.
regression model with nonbinary covariates. Control Clin Trials 2000;21:552–60. 46 Birmingham TB. Test-­retest reliability of lower extremity functional instability
24 Schoenfeld DA. Sample-­size formula for the proportional-­hazards regression model. measures. Clin J Sport Med 2000;10:264–8.
Biometrics 1983;39:499–503. 47 Uchida Y, Demura S. Differences in leg muscle activity and body sway between elderly
25 Therneau TM, Grambsch PM. Modeling survival data: extending the COX model. New adults able and unable to maintain one-­leg stance for 1 min: the effect of hand
York, NY: Springer, 2000: 39–77. support. Aging Clin Exp Res 2016;28:669–77.
26 Harrell FE, Lee KL, Mark DB. Multivariable prognostic models: issues in developing 48 da Silva RA, Vieira ER, Carvalho CE, et al. Age-­related differences on low back
models, evaluating assumptions and adequacy, and measuring and reducing errors. pain and postural control during one-­leg stance: a case-­control study. Eur Spine J
Stat Med 1996;15:361–87. 2016;25:1251–7.
27 DeLong ER, DeLong DM, Clarke-­Pearson DL. Comparing the areas under two or 49 Schorderet C, Hilfiker R, Allet L. The role of the dominant leg while assessing balance
more correlated receiver operating characteristic curves: a nonparametric approach. performance. A systematic review and meta-­analysis. Gait Posture 2021;84:66–78.
Biometrics 1988;44:837–45. 50 Hernán MA. The hazards of hazard ratios. Epidemiology 2010;21:13–15.

Araujo CG, et al. Br J Sports Med 2022;56:975–980. doi:10.1136/bjsports-2021-105360 7 of 7