Group Title: BMC Geriatrics
Title: Physical function and self-rated health status as predictors of mortality: results from longitudinal analysis in the ilSIRENTE study
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00099954/00001
 Material Information
Title: Physical function and self-rated health status as predictors of mortality: results from longitudinal analysis in the ilSIRENTE study
Physical Description: Book
Language: English
Creator: Cesari, Matteo
Onder, Graziano
Zamboni, Valentina
Manini, Todd
Shorr, Ronald
Russo, Andrea
Bernabei, Roberto
Pahor, Marco
Landi, Francesco
Publisher: BMC Geriatrics
Publication Date: 2008
 Notes
Abstract: BACKGROUND:Physical function measures have been shown to predict negative health-related events in older persons, including mortality. These markers of functioning may interact with the self-rated health (SRH) in the prediction of events. Aim of the present study is to compare the predictive value for mortality of measures of physical function and SRH status, and test their possible interactions.METHODS:Data are from 335 older persons aged = 80 years (mean age 85.6 years) enrolled in the "Invecchiamento e Longevità nel Sirente" (ilSIRENTE) study. The predictive values for mortality of 4-meter walk test, Short Physical Performance Battery (SPPB), hand grip strength, Activities of Daily Living (ADL) scale, Instrumental ADL (IADL) scale, and a SRH scale were compared using proportional hazard models. Kaplan-Meier survival curves for mortality and Receiver Operating Characteristic (ROC) curve analyses were also computed to estimate the predictive value of the independent variables of interest for mortality (alone and in combination).RESULTS:During the 24-month follow-up (mean 1.8 years), 71 (21.2%) events occurred in the study sample. All the tested variables were able to significantly predict mortality. No significant interaction was reported between physical function measures and SRH. The SPPB score was the strongest predictor of overall mortality after adjustment for potential confounders (per SD increase; HR 0.64; 95%CI 0.48–0.86). A similar predictive value was showed by the SRH (per SD increase; HR 0.76; 95%CI 0.59–0.97). The chair stand test was the SPPB subtask showing the highest prognostic value.CONCLUSION:All the tested measures are able to predict mortality with different extents, but strongest results were obtained from the SPPB and the SRH. The chair stand test may be as useful as the complete SPPB in estimating the mortality risk.
General Note: Start page 34
General Note: M3: 10.1186/1471-2318-8-34
 Record Information
Bibliographic ID: UF00099954
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: Open Access: http://www.biomedcentral.com/info/about/openaccess/
Resource Identifier: issn - 1471-2318
http://www.biomedcentral.com/1471-2318/8/34

Downloads

This item has the following downloads:

PDF ( PDF )


Full Text



BMC Geriatrics Central



Research article

Physical function and self-rated health status as predictors of
mortality: results from longitudinal analysis in the ilSIRENTE study
Matteo Cesari*1,2, Graziano Onder', Valentina Zamboni1, Todd Manini2,
Ronald I Shorr2,3, Andrea Russo', Roberto Bernabeil, Marco Pahor2 and
Francesco Landi1


Address: 'Dipartimento di Scienze Gerontologiche, Geriatriche e Fisiatriche, Universita Cattolica del Sacro Cuore, Roma, Italy, 2Department of
Aging and Geriatric Research, University of Florida Institute on Aging, Gainesville, FL, USA and 3Veterans Affairs Geriatric Research, Education
and Clinical Center, Gainesville, FL, USA
Email: Matteo Cesari* macesari@gmail.com; Graziano Onder graziano_onder@rm.unicatt.it;
Valentina Zamboni valentina_zamboni@tiscali.it; Todd Manini tmanini@aging.ufl.edu; Ronald I Shorr rshorr@ufl.edu;
Andrea Russo andrea_russo@rm.unicatt.it; Roberto Bernabei roberto_bernabei@rm.unicatt.it; Marco Pahor mpahor@ufl.edu;
Francesco Landi francesco_landi@rm.unicatt.it
* Corresponding author



Published: 22 December 2008 Received: 3 June 2008
BMC Geriatrics 2008, 8:34 doi: 10.1 186/1471-23 18-8-34 Accepted: 22 December 2008
This article is available from: http://www.biomedcentral.com/1471-23 18/8/34
2008 Cesari et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.ore/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.



Abstract
Background: Physical function measures have been shown to predict negative health-related
events in older persons, including mortality. These markers of functioning may interact with the
self-rated health (SRH) in the prediction of events. Aim of the present study is to compare the
predictive value for mortality of measures of physical function and SRH status, and test their
possible interactions.
Methods: Data are from 335 older persons aged > 80 years (mean age 85.6 years) enrolled in the
"Invecchiamento e LongevitA nel Sirente" (ilSIRENTE) study. The predictive values for mortality of
4-meter walk test, Short Physical Performance Battery (SPPB), hand grip strength, Activities of
Daily Living (ADL) scale, Instrumental ADL (IADL) scale, and a SRH scale were compared using
proportional hazard models. Kaplan-Meier survival curves for mortality and Receiver Operating
Characteristic (ROC) curve analyses were also computed to estimate the predictive value of the
independent variables of interest for mortality (alone and in combination).
Results: During the 24-month follow-up (mean 1.8 years), 71 (21.2%) events occurred in the study
sample. All the tested variables were able to significantly predict mortality. No significant
interaction was reported between physical function measures and SRH. The SPPB score was the
strongest predictor of overall mortality after adjustment for potential confounders (per SD
increase; HR 0.64; 95%CI 0.48-0.86). A similar predictive value was showed by the SRH (per SD
increase; HR 0.76; 95%CI 0.59-0.97). The chair stand test was the SPPB subtask showing the
highest prognostic value.
Conclusion: All the tested measures are able to predict mortality with different extents, but
strongest results were obtained from the SPPB and the SRH. The chair stand test may be as useful
as the complete SPPB in estimating the mortality risk.



Page 1 of 9
(page number not for citation purposes)







http://www.biomedcentral.com/1471-2318/8/34


Background
Over the past two decades, there has been a growing rec-
ognition of the functional status assessment as a key factor
in the evaluation of older persons[ 1]. This importance is
mainly due to the large evidence that physical function
measures are not only associated with clinical and sub-
clinical age-related modifications[2,3], but are also able to
predict future health-related events, including disabil-
ity[4,5], institutionalization [6-8], and mortality[6,9].

Among the different possible options to evaluate the
physical functioning of older persons, the use of specifi-
cally designed questionnaires aimed at evaluating how
the subject interacts with the surrounding environment
and identifying his impairments is one of the more com-
monly chosen. Best examples of this kind of tests are the
Activities of Daily Living (ADL)[10], and Instrumental
Activities of Daily Living (IADL)[11] scales, which were
designed about 30-40 years ago. More recently, objective
measures of physical performance and muscle strength
have been developed to obtain objective estimates of the
older persons' physical function. It has been shown that
physical performance and muscle strength measures are
able to identify more limitations in physical functioning
than self-reported or subjective measures[ 12,13], and may
be more useful for longitudinal evaluations because more
sensible to changes[14]. Among these objective measures
of physical function, the Short Physical Performance Bat-
tery (SPPB)[4], the 4-meter walking test[5], and the hand
grip strength[ 15] are the most commonly used in clinical
as well as research settings. Unfortunately, despite of the
demonstrated critical role of physical function in the eval-
uation of older persons, the screening visit of an older per-
son still mostly relies on self-reported questions (mainly
due to time and/or space restrains commonly present
especially in the clinical practice).

Similarly to physical function measures, self-rated health
(SRH) has been shown to significantly predict negative
outcomes (including disability[16] and mortality [17-
20]). It has been explained that SRH might better capture
the burden of clinical and subclinical conditions com-
pared to the traditionally adopted measures of disease, or
that positive self-ratings may mirror a general optimistic
disposition[2 1] (consequently promoting a virtuous cycle
with beneficial effects on neurological, immunological
and endocrinological pathways[22]). The relationship
between SRH and mortality has recently been shown to be
independent of several potential confounders, including
physical function (i.e., isometric muscle strength)[22].

However, some issues about the use of all these measures
are still present and need clarification. Firstly, although all
of the above-presented measures have shown to be predic-
tive of mortality, a direct comparison among them for this


outcome has not yet been formally conducted in litera-
ture. Secondly, physical performance and muscle strength
measures (which on a broader extent represent markers of
well-being) may interact with the self-perception of the
health status in the prediction of events, but this hypoth-
esis has never been explored. Investigating it may provide
useful insights on the best way to manage all these screen-
ing instruments. It is noteworthy that evidence is particu-
larly lacking for the very old persons. This aged group is
the one in which evidence-based medicine is often very
difficult to apply and clinical decisions are often driven by
the subject's feelings. Therefore, all the screening instru-
ments we plan to examine play a major rule in the deter-
mination of the frail older individual physical, functional,
and biological reserves. Finally, the age-related decline in
physical function due to the higher number of clinical and
subclinical conditions may modify the predictive value of
the commonly used measures of physical function as well
as the self-perceived health status in the oldest old. Con-
sequently, the clinical meaning of these markers may
change in this age group.

In the present study, we hypothesized that 1) physical
function and self-rated health (SRH) measures are predic-
tive of negative health-related events in very old persons,
and 2) a possible added effect of these instruments may
allow a better prediction of events compared to when a
single test is used. Therefore, we compared the predictive
value for mortality of several measures of physical func-
tion (i.e. two measures of physical performance, the
SPPB[4] and the 4-meter walking speed test; a marker of
muscle strength, the hand grip strength; and two scales of
disability, the ADL[10] and the IADL[11] scales), and a
self-perceived measure of well-being (i.e. a SRH scale[23])
in a sample of very old persons (aged 80 years and older)
enrolled in the "Invecchiamento e Longevita nel Sirente"
(Aging and longevity in the Sirente geographic area, ilSI-
RENTE) study[24].

Methods
We used baseline data from the ilSIRENTE, a prospective
cohort study performed in the mountain community liv-
ing in the Sirente geographic area (L'Aquila, Italy) and
developed by the teaching nursing home Opera Santa
Maria della Pace (Fontecchio, L'Aquila, Italy) in a partner-
ship with local administrators and primary care physi-
cians. Details of the design and methods of ilSIRENTE
have been described elsewhere[24]. Briefly, potential
study participants were identified by selecting from the
Registry Offices every person born before 1st January 1924
and still living in the municipalities involved in the study
at the end of October 2003. A total of 364 participants
were enrolled in the study. Participants' baseline assess-
ments began in December 2003 and were completed in
September 2004. Clinical interview and functional assess-


Page 2 of 9
(page number not for citation purposes)


BMC Geriatrics 2008, 8:34







http://www.biomedcentral.com/1471-2318/8/34


ment were performed at the study clinics located in each
town. Home visit was performed if participant was unable
to reach the study clinic. Information was obtained by the
participant or, if he/she was incapable, by a proxy. The
University Cattolica del Sacro Cuore (Roma, Italy) Ethical
Committee approved the study protocol. All the partici-
pants signed an informed consent at the baseline visit.

The present analyses were conducted in 335 participants,
after exclusion of 29 participants with missing data for the
main variables of interest.

The Minimum Data Set for Home Care (MDS-HC)
The Minimum Data Set for Home Care (MDS-HC) instru-
ment[251 was administered to all study participants. The
MDS-HC contains a variety of different, multi-item sum-
mary scales, exploring socio-demographics, clinical diag-
noses, and physical function status. Besides, the MDS-HC
includes information about an extensive array of signs,
symptoms, syndromes, and treatments. The MDS items
have shown an excellent inter-rater and test-retest reliabil-
ity when completed by nurses performing usual assess-
ment duties (average weighted Kappa = 0.8[26,27]). A
questionnaire exploring family history, lifestyle, nutri-
tion, physical activity, and other behavioral factors shared
with the "Invecchiare in Chianti" (Aging in the Chianti
geographic area, InCHIANTI) study[281 was also addi-
tionally administered.

Mortality
Vital status of all the study participants was ascertained
from the general practitioners, and confirmed by the
National Death Registry until 24 months after the base-
line visit. The follow-up time considered for the present
analyses was calculated as the time from the date of base-
line visit to the date of death (for participants who died
during the follow-up), and censored to 24 months for par-
ticipants who did not die during the study follow-up.

Physical performance, muscle strength and functional
status measures
Physical performance measures
Physical performance was assessed by the 4-meter walking
speed and the Short Physical Performance Battery score.
This latter measure is composed of three timed tests: 4-
meter walking speed, balance, and chair stand tests[5].
Timed results from each test were categorized into 5-level
variables ranging from 0 (worst performers) to 4 (best per-
formers) according to well-established cut-points[51. The
sum of the results from the three categorized tests (rang-
ing from 0 to 12) was used for the present analyses.

Walking speed was evaluated measuring the participant's
usual gait speed (in m/sec) over a 4-meter course. The fol-
lowing cut-points were used to categorize the variable:


<0.46 m/s, a score of 1; 0.46 to 0.64 m/s, a score of 2; 0.65
to 0.82 m/s, a score of 3; > 0.83 m/s, a score of 4. Partici-
pants unable to complete the task were scored 0.

To assess the chair stand test, participants were asked to
stand up from a chair with their arms folded across the
chest five times in a row as quickly as possible. The time
needed to complete the task was recorded. The following
cut-points were used to categorize the variable:> 16.7 sec-
onds, a score of 1; 13.7 to 16.6 seconds, a score of 2; 13.6
to 11.2 seconds, a score of 3; and < 11.1 seconds, a score
of 4. Subjects unable to complete the test received a score
of 0.

To assess the balance test, participants were asked to per-
form three increasingly challenging standing positions:
side-by-side position, semi-tandem position, and tandem
position. Participants were asked to hold each position for
10 seconds. Participants were scored as 1 if they were able
to hold a side-by-side standing position for 10 seconds,
but were unable to hold a semi-tandem position for 10
seconds; a score of 2 if they were able to hold a semi-tan-
dem position for 10 seconds, but were unable to hold a
tandem position for more than 2 seconds; a score of 3 if
they were able to stand in tandem position for 3 to 9 sec-
onds; and a score of 4 if they were able to hold the tandem
position for 10 seconds. Participants unable to complete
the test were scored 0.

Muscle strength measure
Muscle strength was assessed by hand grip strength meas-
ured by a dynamometer (North Coast Hydraulic Hand
Dynamometer, North Coast Medical Inc, Morgan Hill,
CA, USA). One trial for each hand was performed, and the
result from the strongest hand was used in the present
analyses. Hand grip strength has shown to be predictive of
major health-related events in older persons[ 15,29].

Functional status measures
In the ilSIRENTE study, Basic and Instrumental Activities
of Daily Living (ADL and IADL, respectively) scales were
assessed as part of the MDS-HC instrument[26]. The
assessor evaluated the participants' capacity to perform
each task included in the ADL and IADL scales. Being the
MDS-HC a comprehensive geriatric assessment tool
aimed at 1) identifying the critical issues of the health sta-
tus and care of older persons and 2) designing a specifi-
cally-tailored intervention plan, the impairment in each
task was defined as the disability and/or the need of assist-
ance in adequately performing the task. Therefore, all the
ADL and IADL items were coded as "0" if the participant
was independent in performing the specific task, or as "1"
if supervision was required and/or the participant was
completely dependent. The ADL scale (range 0-7, a higher
number indicates higher impairment) is composed by the


Page 3 of 9
(page number not for citation purposes)


BMC Geriatrics 2008, 8:34







http://www.biomedcentral.com/1471-2318/8/34


following tasks: eating, dressing, personal hygiene, mobil-
ity in bed, dressing, transferring (from/to bed, chair or
stand position), use of the toilet. The IADL scale (range 0-
7, a higher number indicates higher impairment)
included: meal preparation, shopping, telephone use,
housekeeping, responsibility for medication intake, han-
dling finances, use of transportation.

Self-assessed health status
At the baseline visit, a single-item measure of SRH was
administered to all the participants[23]. Subjects were
asked to answer to the question "How is your health in
general?" rating their status as "Very Poor", "Poor",
"Sometimes Good, Sometimes Poor", "Good", or "Very
Good". The relative score ranging from 1 (worst) to 5
(best) was used for the present analyses. SRH has shown
to be a powerful predictor of mortality[ 17].

Covariates
Covariates considered in the present analyses include:
sociodemographic characteristics (age, gender, and smok-
ing habit), body mass index (BMI), comorbidity, medica-
tions, and biological marker levels (albumin and total
cholesterol). Body weight was measured with light clothes
using a calibrated scale. Body height was measured using
a standard stadiometer. BMI was defined as weight (in kil-
ograms) divided by the square of height (in meters). The
cognitive performance scale (CPS) was used to assess cog-
nitive status[30]. The CPS has shown an excellent inter-
rater and test-retest reliability when completed by nurses
performing usual assessment duties[26]. The CPS score
ranges from 0 (intact cognition) to 6 (severely impaired
cognition). The following clinical diagnoses were assessed
by a study physician on the basis of self- (or proxy-)
reported history and clinical records review and consid-
ered in the adjusted analyses: coronary heart disease, con-
gestive heart failure, cerebrovascular disease, diabetes,
cancer, depression, dementia. A cumulative index of
comorbidity defined by the number of clinical conditions
was used for the present analyses. Standard determina-
tions of total cholesterol and serum albumin concentra-
tions were determined by using commercial available
kits suitable for use on Olympus 2700 instrumentation
(Olympus, Milano, Italy). C-Reactive protein (CRP) con-
centrations were determined by a high sensitivity Enzyme-
Linked ImmunoSorbent Assay kit (Bender MedSystems,
Vienna, Austria). The CRP assay had a sensitivity of 3 pg/
mL. The intra-assay coefficient of variation was 6.9%.

Statistical analysis
Means (and standard deviations, SD), proportions (in
percentage) were calculated to describe the main charac-
teristics of the study sample. Unadjusted and adjusted
proportional hazard models were performed to estimate
the hazard ratios (HR, and 95% confidence intervals,


95%CI) of mortality (dependent variable) for physical
performance, muscle strength, physical function, and SRH
variables (independent variables). To permit direct com-
parisons of predictors, all the results are shown per SD
increase of the measures. Kaplan-Meier survival curves for
mortality were also performed according to physical per-
formance and SRH groups. Receiver Operating Character-
istic (ROC) curve analyses were also computed to estimate
the predictive value of the independent variables of inter-
est for mortality (alone and in combination) through the
evaluation of the Areas Under the Curves (AUCs). A p
value < 0.05 was chosen for statistical significance for all
the present analyses. All the analyses were performed
using SPSS software (version 13.0, SPSS Inc., Chicago, IL).

Results
Main characteristics of the study sample population (n =
335; mean age 85.6 [SD 4.8] years) are presented in Table
1 according to vital status at the end of follow-up (mean
length 1.8 [SD 0.5] years; 71 [21.2%] events). Compared
to participants alive at the end of the follow-up, those
who died were older and had a higher prevalence of con-
gestive heart failure, cerebrovscular disease, depression,
diabetes, and dementia. They also had lower BMI, albu-
min, and total cholesterol, and higher CRP concentration.
For what concern the variables of interest for the present
study, all the measures of physical function as well as the
SRH score were significantly higher in participants alive at
the end of the follow-up compared to cases.

Results from unadjusted and adjusted proportional haz-
ard models predicting mortality for all the variables of
interest (per their SD increases) are shown in Table 2.
Inverse and significant relationships of all the measures of
physical function and SRH with mortality were found,
even when models were adjusted for age and gender (all p
values < 0.001). However, when additional potential con-
founders (i.e. number of clinical conditions and biologi-
cal markers) were included into the models, the hand grip
strength and the ADL score lost their statistical signifi-
cance. In the final adjusted model, when also CRP con-
centration (log value) was included as covariate, the only
SPPB and the SRH scores maintained their statistical sig-
nificance in their association with mortality (HR 0.64,
95%CI 0.48-0.86; p = 0.003, and HR 0.76, 95%CI 0.59-
0.97; p = 0.03, respectively).

Separate partially adjusted models (for statistical power
reasons; Model 1 adjustment) were also performed using
the categorical variables (ranging from 0 to 4) for each
subtask of the SPPB as independent variable of interest.
Statistically significant and positive associations were
reported between all the SPPB tasks and survival (all p for
trend < 0.001). At the 4-meter walking speed test, partici-
pants scoring 1, 2, 3, and 4 had 56.3%, 63.6%, 83.1%,


Page 4 of 9
(page number not for citation purposes)


BMC Geriatrics 2008, 8:34








http://www.biomedcentral.com/1471-2318/8/34


Table I: Main characteristics (mean standard deviation, o
percentage) of the study sample according to mortality.


Death No death
(n = 71) (n = 264)


Sociodemographic
Age (years)
Gender (women)
Education (years)
Current smoking
Body Mass Index (kg/m2)
Cognitive Performance Scale

Clinical conditions
Coronary artery disease
Congestive heart failure
Cerebrovascular disease
Cancer
Depression
Diabetes
Dementia
Number of clinical conditions

Biological markers
Albumin (g/dL)
Total cholesterol (mg/dL)
C-reactive protein (mg/L)

Physical performance measures
4-meter walking speed
SPPB score (0-12)
Walk test (0-4)
Chair stand test (0-4)
Balance test (0-4)
Hand grip strength (kg)
Functional status measures
ADL score
IADL score
Self-rated health status
Self-Rated Health score


SPPB: Short Physical Performanc
Daily Living (range 0 [totally dep
IADL: Instrumental Activities of
dependent] 7 [totally independ
Cognitive Performance Scale ran
Health score range: I (worst) -

and 85.7% lower mortality
ErouD (i.e. subjects scoring


87.5 5.2
64.8
4.9 1.1
2.8
24.4 4.2
1.6 1.9


15.5
15.5
14.1
8.5
35.2
36.6
14.1
1.4 1.2


85.0 4.6
67.0
5.2 1.8
2.3
26.1 4.5
0.6 1.3


S1.0
3.4
1.9
3.4
23.9
25.0
5.3
0.7 0.8


r


found between physical function measures and SRH. No


significant gender interaction was found between all the
p independent variables of interest and mortality (all p val-
ues for interaction terms >0.3).


<0.001
0.72
0.14
0.79
0.003
<0.001


0.30
<0.001
<0.001
0.07
0.05
0.05
0.01
<0.00 I


Since the chair stand test was the SPPB subtask showing
the strongest association with mortality, secondary analy-
ses were performed to evaluate the possible existence of
an additive predictive value for mortality of this SPPB
component with the SRH.

No significant differences were found among AUCs
designed by ROC curve analyses for mortality when the
chair stand test (AUC 0.725, 95%CI 0.661-0.789), the
SRH (AUC 0.656, 95%CI 0.582-0.730), and their combi-
nation (AUC 0.751, 95%CI 0.686-0.816) were tested.
Similar findings were reported when testing the SPPB
score (AUC 0.743, 95%CI 0.679-0.806), and its combi-
nation with the SRH (AUC 0.749, 95%CI 0.683-0.814).


The chair stand test was then categorized according to the

4.1 0.4 4.2 0.3 0.001 ability (n = 226, 67.5%) or not (n = 109, 32.5%) to per-
178.7 38.0 203.6 45.2 <0.001 form the task. The SRH score was categorized according to
5.1 (2.5-7.4) 2.5 (1.3-4.6) <0.001 the median value in two groups (i.e. SRH score 170 [47.9%]; SRH score >3: n = 165 [52.1%]). Figure 1
shows results from Kaplan-Meier survival curves for mor-
0.35 0.30 0.55 0.28 <0.001 tality according to physical performance and SRH groups.
4.0 3.3 7.1 3.3 <0.001 Participants able to complete the chair stand test were sig-
1.3 1.1 2.1 1.2 <0.00 nificantly less likely to die compared to those with poor

2.0 1.6 3.1 1.3 <0.001 physical performance (p < 0.001). No significant differ-
27.7 14.4 33.2 13.7 <0.00 I ences were found 1) among participants able to complete
the chair stand test, or 2) among participants unable to
4.3 2.2 6.2 1.8 <0.001 complete the chair stand test, according to the SRH status
2.4 2.5 4.7 2.3 <0.001 groups (pairwaise comparisons p = 0.47, and p = 0.17,
respectively). An adjusted multivariable proportional haz-
2.97 0.89 3.48 0.83 <0.001 ard model (Table 3) confirmed these findings, showing
e Battery; ADL: Basic Activities of that participants unable to complete the chair stand test
e Battery; ADL: Basic Activities of
endent] 7 [totally independent]); and with worse SRH had a higher risk of mortality (HR
Daily Living (range 0 [totally 2.36, 95%CI 1.16-4.79; p = 0.02) compared to the refer-
ent]). ence group (i.e. able to complete the chair stand test and
ge: 0 (best) 6 (worst); Self-Rated SRH score >3), even after adjustment for all the potential
5 (best). confounders. Consistent results were obtained when the

risk compared to the reference overall SPPB score was tested in combination with the
S0). resnectivelv. Similar find- SRH.


ings were also found for the chair stand (45.9%, 72.8%,
92.5%, and 86.6%, respectively) and the balance (47.0%,
58.0%, 69.7%, and 79.2%, respectively) tests.

Analyses were also conducted to evaluate which of the
three subtasks composing the SPPB score was the most
strongly associated with mortality. Results showed that
the chair stand test was the only significantly associated
with mortality, while only borderline significance were
found for the balance, and the 4-meter walk test. No sig-
nificant interaction for the prediction of mortality was


Discussion
In the present study, we compared the predictive value for
mortality of two physical performance measures (i.e. 4-
meter walk test, and the SPPB score), a measure of muscle
strength (i.e. hand grip), two measures of disability (i.e.
ADL and IADL scores) and a self-reported measure of
well-being (i.e. a SRH scale). Our results from unadjusted
and partially adjusted analyses showed that all the tested
physical function variables were able to predict mortality.
However, the SPPB score was the strongest predictor of



Page 5 of 9
(page number not for citation purposes)


BMC Geriatrics 2008, 8:34








http://www.biomedcentral.com/1471-2318/8/34


Table 2: Proportional hazard models between measures of health status (per standard deviation increase) and mortality.


Unadjusted
HR (95%CI)


Model I
HR (95%CI)


Model 2
HR (95%CI)


Model 3
HR (95%CI)


Objective measures
4-meter walking speed
SPPB score
SPPB Walk test
SPPB Balance test
SPPB Chair stand test
Hand grip strength (kg)
ADL score
IADL score
SRH score


0.49 (0.38-0.63)t
0.44 (0.34-0.57)t
0.48 (0.36-0.64)t
0.55 (0.44-0.68)t
0.38 (0.27-0.53)t
0.55 (0.43-0.71)t
0.54 (0.45-0.65)t
0.48 (0.38-0.61)t
0.62 (0.50-0.77)t


0.53 (0.40-0.69)t
0.46 (0.36-0.60)t
0.52 (0.39-0.70)t
0.58 (0.47-0.74)t
0.40 (0.29-0.56)t
0.57 (0.44-0.75)t
0.58 (0.48-0.71)t
0.51 (0.40-0.67)t
0.65 (0.52-0.80)t


0.73 (0.54-0.99)*
0.62 (0.47-0.83)t
0.71 (0.52-0.97)*
0.77 (0.60-1.00)
0.51 (0.36-0.72)t
0.77 (0.58-1.04)
0.77 (0.59-1.00)
0.70 (0.50-0.99)*
0.78 (0.60-0.99)*


0.77 (0.56-1.05)
0.64 (0.48-0.86)t
0.73 (0.54-1.01)
0.78 (0.60-1.01)
0.54 (0.38-0.76)t
0.78 (0.58-1.05)
0.81 (0.62-1.06)
0.77 (0.54-1.09)
0.76 (0.59-0.97)*


* p < 0.05; p < 0.01
SPPB: Short Physical Performance Battery; ADL: Basic Activities of Daily Living; IADL: Instrumental Activities of Daily Living; SAHS: Self-assessed
health status.
Standard deviations: 4-meter walking speed 0.29714; SPPB score 3.50676; SPPB Walk test score 1.22735; SPPB Balance test score 1.45514; SPPB
Chair stand test score 1.42872; Hand grip strength 14.28956 kg; Basic Activities of Daily Living 2.19641; Instrumental Activities of Daily Living
2.51375; Self-rated health 0.86923
Model 1: Adjusted for age and gender.
Model 2: Adjusted for Model I + body mass index, cognitive performance scale, number of clinical conditions, albumin, total cholesterol.
Model 3: Adjusted for Model 2 + C-reactive protein (log value)


overall mortality in these very old community-dwelling
subjects, even after considering several socio-demo-
graphic, clinical, and biological confounders. A lower, but
still significant, predictive value was only showed by the
SRH measure. Among the three subtasks of the SPPB, the
chair stand test was the one showing the highest prognos-
tic value. The combination of the chair stand test and the
SRH score did not provide significant additional benefits
in predicting mortality. In fact, participants with a good
physical performance had a lower risk of dying compared
to those with poor performance, independently of their
self-perceived health status. Moreover, no significant dif-
ferences were reported when comparing the AUC
designed by ROC curve analyses for mortality. However,
when the chair stand test and the SRH results were com-
bined, the selection of the participants poorly scoring at
both tests, led to the identification of a smaller number of
subjects characterized by the highest risk of mortality
compared to participants with good physical performance
and SRH.


Previous studies have already explored the relationship
existing between objective and self-reported measures of
physical function for major health-related events[31,32].
Moreover, the strong relationship between physical per-
formance tests and negative health-related outcomes in
the elderly has already been documented[4,6,33,34].
However, besides of being confirmatory of previous find-
ings showing the importance of physical performance in
older persons, our study still adds some novel contribu-
tions to the topic.

To our knowledge, a direct comparison of the predictive
value for mortality of different screening instruments
(particularly aimed at the evaluation of the physical func-
tion and the health status) is not yet available in literature,
especially among very old subjects. This comparison led
to the identification of the SPPB (in particular, of the chair
stand test subtask) and the SRH as the best predictors of
mortality. It is noteworthy that both these instruments are
quick and inexpensive measures whose implementation


Table 3: Results from a single multivariable proportional hazard model* exploring the relationship of physical performance and self-
rated health with mortality.


NIN (%)


Hazard Ratio (95% Confidence Interval)


SPPB Chair stand test > 0, SRH > 3
SPPB Chair stand test > 0, SRH < 3
SPPB Chair stand test = 0, SRH >3
SPPB Chair stand test = 0, SRH < 3


14/133 (10.5)
14/93 (15.1)
9/32 (28.1)
34/77 (44.2)


I (Reference group)
1.02 (0.46-2.25)
1.45 (0.57-3.74)
2.36 (1.16-4.79)


n/N: number of events/number of participants; SPPB: Short Physical Performance Battery; SRH: Self-rated health
*Adjusted for age, gender, body mass index, cognitive performance scale, number of clinical conditions, albumin, total cholesterol, C-reactive
protein (log value)


Page 6 of 9
(page number not for citation purposes)


BMC Geriatrics 2008, 8:34







http://www.biomedcentral.com/1471-2318/8/34


1.0

0.8

0.6 -

0.4

0.2

0


--- ,-- .



Log rank p
- Able to perform chair stand test; SRH "Good" or better
-- Able to perform chair stand test; SRH worse than "Good"
........ Unable to perform chair stand test; SRH "Good" or better
- -- Unable to perform chair stand test; SRH worse than "Good"


0 0.5 1.0 1.5
Follow-up (years)


Figure I
Kaplan-Meier survival curves for mortality according
to the ability to perform the chair stand test and self-
rated health (SRH) score.


in clinical settings may not be particularly honerous in
terms of training, costs, and time.

Our results did not show evidence of a statistical interac-
tion between SPPB, SRH and mortality. However, the sur-
vival analyses we performed seem to suggest such
interaction. In fact, the SPPB score was able to discrimi-
nate participants with higher mortality risk regardless of
their SRH status. This was not evident for the SRH instru-
ment, which tended to discriminate individuals at higher
risk only among those with poor physical performance.
This finding may support the use of SPPB (and physical
performance measures in general) as optimal screening
tools for older persons, independently of their health sta-
tus.

Our analyses of the three subtests composing the SPPB
demonstrated that the chair stand test was more strongly
associated with the mortality outcome than the walk and
balance tests. Moreover, the chair stand test was able to
predict mortality in a very similar way to the complete
SPPB score. This may suggest that the adoption of this
only subtask in those settings with time and/or space
restrains might already be sufficient to identify older per-
sons at risk of events. In the attempt to facilitate the pos-
sible implementation of this test in the clinical setting, our
secondary analyses tested a dichotomous variable of the
chair stand test defined as the ability or not to stand up
from the chair five times in a row. The adoption of this
single SPPB subtask as screening tool for older persons
may be very easy to implement, even more than a walking
speed test. Interestingly, Ensrud and colleagues recently
proposed a frailty index including the inability to rise


from a chair 5 times without using arms as a component
criterion[35]. Authors compared the predictive value of
this new index to that of the more commonly used (but
more complex) Fried and colleagues' one[36], reporting
similar results. Our results showing the higher prognostic
value of the chair stand test in comparison with the other
SPPB subtasks is not completely in line with the sparse
previous evidence. In fact, the few studies available on the
topic suggest that the walking speed is the most sensitive
subtask of the SPPB in predicting incident disabil-
ity[4,37], and mortality[38]. A possible explanation to the
different results we found might be the older age of our
sample population. It might be that the three subtests
composing the SPPB may present different age-related
declines. If the 4-meter walk test is more prematurely
affected by aging (and the related underlying conditions),
a "floor" effect may limit the predictive value of it in favor
of a possible more stable test (i.e. chair stand test).

The predictive value for mortality of SRH, independently
of health risk factors is well-demonstrated in litera-
ture[17]. Several explanations to this relationship have
been provided. It is possible that SRH may better capture
the burden of diseases and symptoms. Another explana-
tion might be related to the wider spectrum of informa-
tion (inclusive of personal sensations) that a person can
describe when self-reporting the own health status, and
which may partially be excluded by "external" evalua-
tions[39].

Significant results in the prediction of mortality were
reported by the hand grip strength, and the IADL andADL
scales only in the unadjusted and partially adjusted mod-
els. It is noteworthy that the predictive values of the hand
grip strength, ADL, and IADL scores for mortality were
strongly weakened by inclusion in the statistical models
of clinical conditions and, later, CRP concentrations. Con-
sistent results have previously been reported in studies
testing the associations of these measures with comorbid-
ity[32], health status[ 14], and incident health-related
events[33,40]. The hand grip strength is a standardized
measure of a specific muscular district strength which is
generalized to the overall individual muscular function-
ing[15,41]. On the other hand, the SPPB requires a good
overall physical functioning of the subject to be success-
fully completed. The ADL and IADL scales are designed to
evaluate the ability of a subject to interact with the sur-
rounding environment and independently accomplish
crucial tasks of life[10,11]. Thomas and colleagues
recently showed that objective measures of physical per-
formance are able to improve the assessment of func-
tional status provided by subjective measures of physical
function in older persons[ 13]. Therefore, it is likely that
the SPPB is able to capture a wider scope of information
from different sources related to physical functioning than


Page 7 of 9
(page number not for citation purposes)


BMC Geriatrics 2008, 8:34








http://www.biomedcentral.com/1471-2318/8/34


hand grip strength, and ADL and IADL scales. In this con-
text, it may also not be surprising that a general SRH meas-
ure is more strongly associated to mortality than
specifically-aimed subjective screening tests (i.e. ADL and
IADL scales) or too sectorial measures (i.e. hand grip
strength).

Our study presents some limitations. The limited sample
size may have influenced some of our results, potentially
leading to type I errors. However, the risk of false negative
results may be limited due to the overall consistency of
our findings (even with previous reports). Our sample
population was composed by older community-dwelling
persons aged 80 years and older. Further studies confirm-
ing our findings, and extending them to different age
groups, settings and populations are needed. Third factors
not considered in our study (e.g. body composition), and
potentially explaining (at least partly) our results may rep-
resent a further limit of the present analyses.

Conclusion
Our study shows that all the tested measured are able to
predict mortality with different extents. However, the only
which are not influenced by sociodemographic, clinical,
and biological factors in their prediction are the SPPB and
the SRH. The chair stand test may be as useful as the com-
plete SPPB in estimating the mortality risk, and the testing
of the only ability to perform it may already be sufficient
to provide useful prognostic information.

Competing interests
The authors declare that they have no competing interests.

Authors' contributions
MC conceived and designed the study, carried out the data
analysis and interpreted the results, drafted the manu-
script, and gave the final approval. GO conceived and
designed the study, carried out the data analysis and inter-
preted the results, helped in drafting the manuscript, pro-
vided critical review of the manuscript, and gave the final
approval. VZ conceived and designed the study, was
involved in the data acquisition, helped in the interpreta-
tion of the data, provided critical review of the manu-
script, and gave the final approval. TM helped in the
interpretation of the data, provided critical review of the
manuscript, helped in drafting the manuscript, and gave
the final approval. RIS helped in the interpretation of the
data, provided critical review of the manuscript, and gave
the final approval. AR conducted the acquisition of the
data, provided critical review of the manuscript, and gave
the final approval. RB helped in the interpretation of the
data, provided critical review of the manuscript, and gave
the final approval. MP was involved in the conception and
design of the study, helped in the interpretation of the
data, provided critical review of the manuscript, and gave


the final approval. FL conceived and designed the study,
was responsible for the acquisition of the data, helped in
the analysis and interpretation of the data, provided criti-
cal review of the manuscript, and gave final approval.

Acknowledgements
The "Invecchiamento e Longevita' nel Sirente" (ilSIRENTE) study was sup-
ported by a grant from the "Comunita' Montana Sirentina" (Secinaro,
L'Aquila, Italy). Drs. Cesari and Pahor are supported by the University of
Florida Institute on Aging and the Claude D. Pepper Older Americans Inde-
pendence Center (NIH grant I P30AG028740).

References
I. Applegate WB, Blass JP, Williams TF: Instruments for the func-
tional assessment of older patients. N Engi J Med 1990,
322:1207-1214.
2. St-Onge MP: Relationship between body composition changes
and changes in physical function and metabolic risk factors in
aging. Curr Opin Clin Nutr Metab Care 2005, 8:523-528.
3. GuralnikJM, Branch LG, Cummings SR, Curb JD: Physical perform-
ance measures in aging research. J Gerontol 1989,
44:M141-M146.
4. Guralnik JM, Ferrucci L, Simonsick EM, Salive ME, Wallace RB:
Lower-extremity function in persons over the age of 70 years
as a predictor of subsequent disability. N EngI J Med 1995,
332:556-561.
5. Guralnik JM, Ferrucci L, Pieper CF, Leveille SG, Markides KS, Ostir
GV, et al.: Lower extremity function and subsequent disability:
consistency across studies, predictive models, and value of
gait speed alone compared with the Short Physical Perform-
ance Battery. J Gerontol A Biol Sci Med Sci 2000, 55(4):M221 -M23 I.
6. GuralnikJM, Simonsick EM, Ferrucci L, Glynn RJ, Berkman LF, Blazer
DG, et al.: A short physical performance battery assessing
lower extremity function: association with self-reported dis-
ability and prediction of mortality and nursing home admis-
sion. J Gerontol 1994, 49:M85-M94.
7. Studenski S, Perera S, Wallace D, Chandler JM, Duncan PW, Rooney
E, et al: Physical performance measures in the clinical setting.
J Am Geriatr Soc 2003, 51:314-322.
8. Fogel JF, Hyman RB, Rock B, Wolf-Klein G: Predictors of hospital
length of stay and nursing home placement in an elderly
medical population. j Am Med Dir Assoc 2000, 1:202-210.
9. Inouye SK, Peduzzi PN, Robinson JT, Hughes JS, Horwitz JS, Horwitz
RI, et aL.: Importance of functional measures in predicting
mortality among older hospitalized patients. JAMA 1998,
279:1187-1193.
10. Katz S, Ford AB, Moskowitz RW, Jackson BA, Jaffe MW: Studies of
illness in the aged. The index of ADL: a standardized meas-
ure of biological and psychosocial function. JAMA 1963,
185:914-919.
I I. Lawton MP, Brody EM: Assessment of older people: self-main-
taining and Instrumental Activities of Daily Living. Gerontolo-
gist 1969, 9:179-186.
12. Brach JS, VanSwearingen JM, Newman AB, Kriska AM: Identifying
early decline of physical function in community-dwelling
older women: performance-based and self-report measures.
Phys Ther 2002, 82:320-328.
13. Thomas DR, Marren K, Banks W, MorleyJ: Do objective measure-
ments of physical function in ambulatory nursing home
women improve assessment of functional status? JAm Med Dir
Assoc 2007, 8:469-476.
14. Rozzini R, Frisoni GB, Bianchetti A, Zanetti 0, Trabucchi M: Physical
Performance Test and Activities of Daily Living scales in the
assessment of health status in elderly people. j Am Geriatr Soc
1993, 41:1109-1113.
15. Rantanen T, GuralnikJM, Foley D, Masaki K, Leveille SG, Curb JD, et
al.: Midlife hand grip strength as a predictor of old age disa-
bility. JAMA 1999, 281:558-560.
16. Tas U, Verhagen AP, Bierma-Zeinstra SM, Odding E, Koes BW: Prog-
nostic factors of disability in older people: a systematic
review. BrJ Gen Pract 2007, 57:3 19-323.



Page 8 of 9
(page number not for citation purposes)


BMC Geriatrics 2008, 8:34








http://www.biomedcentral.com/1471-2318/8/34


17. Idler EL, Benyamini Y: Self-rated health and mortality: a review
of twenty-seven community studies. J Health Soc Behav 1997,
38:21-37.
18. MosseyJM, Shapiro E: Self-rated health: a predictor of mortal-
ity among the elderly. Am J Public Health 1982, 72:800-808.
19. Benyamini Y, Blumstein T, Lusky A, Modan B: Gender differences
in the self-rated health-mortality association: is it poor self-
rated health that predicts mortality or excellent self-rated
health that predicts survival? Gerontologist 2003, 43:396-405.
20. Kaplan GA, Camacho T: Perceived health and mortality: a nine-
year follow-up of the human population laboratory cohort.
Am J Epidemiol 1983, I 17:292-304.
21. Benyamini Y, Idler EL, Leventhal H, Leventhal EA: Positive affect
and function as influence on self-assessments of health:
expanding our view beyond illness and disability. j Gerontol B
Psychol Sci Soc Sci 2000, 55:P 07-PI 16.
22. Lyyra TM, Heikkinen E, Lyyra AL, Jylha M: Self-rated health and
mortality: could clinical and performance-based measured
of health and functioning explain the association? Arch Geron-
tol Geriatr 2006, 42:277-288.
23. Simon JG, De Boer JB, Joung IMA, Bosma H, Mackenbach JP: How is
your health in general? A qualitative study on self-assessed
health. Eur] Publ Health 2005, 15:200-208.
24. Landi F, Russo A, Cesari M, Barillaro C, Onder G, Zamboni V, et al.:
The ilSIRENTE study: a prospective cohort study on persons
aged 80 years and older living in a mountain community of
Central Italy. Aging Clin Exp Res 2005, 17:486-493.
25. Morris JN, Fries BE, Bernabei R, Ikegami N, Gilgen R, Steel K, et al.:
RAI Home Care assessment manual Washington, DC: InterRAI Copo-
ration; 1996.
26. Hawes C, Morris JN, Phillips CD, Mor V, Fries BE, Nonemaker S:
Reliability estimates for the Minimum Data Set for nursing
home resident assessment and care screening (MDS). Geron-
tologist 1995, 35:172-178.
27. Landi F, Tua E, Onder G, Carrara B, Sgadari A, Rinaldi C, et al.: Min-
imum data set for home care: a valid instrument to assess
frail older people living in the community. Med Care 2000,
38:1184-1190.
28. Ferrucci L, Bandinelli S, Benvenuti E, Di lorio A, Macchi C, Harris TB,
et al.: Subsystems contributing to the decline in ability to
walk: bridging the gap between epidemiology and geriatric
practice in the InCHIANTI study. j Am Geriatr Soc 2000,
48:1618-1625.
29. Al Snih S, Markides KS, Ray L, Ostir GV, Goodwin JS: Handgrip
strength and mortality in older Mexican Americans. jAm Ger-
iatr Soc 2002, 50:1250-1256.
30. Morris JN, Fries BE, Steel K, Ikegami N, Bernabei R, Carpenter GI, et
al.: Comprehensive clinical assessment in community setting:
applicability of the MDS-HC. J Am Geriatr Soc 1997,
45:1017-1024.
31. GuralnikJM, Simonsick EM, Ferrucci L, Glynn RJ, Berkman LF, Blazer
DG, et al.: A short physical performance battery assessing
lower extremity function: association with self-reported dis-
ability and prediction of mortality and nursing home admis-
sion. J Gerontol 1994, 49:M85-M94.
32. Cesari M, Onder G, Russo A, Zamboni V, Barillaro C, Ferrucci L, et
al.: Comorbidity and physical function Results from the ilSI-
RENTE study. Gerontology 2005, 52:24-32.
33. Onder G, Penninx BWJH, Lapuerta P, Fried LP, Ostir GV, Guralnik
JM, et al.: Change in physical performance over time in older
women: the Women's Health and Aging Study. j Gerontol A
Biol Sci Med Sci 2002, 57A:M289-M293.
34. Cesari M, Kritchevsky SB, Penninx BWJH, Nicklas BJ, Simonsick EM,
Newman AB, et al.: Prognostic value of usual gait speed in well-
functioning elders Results from the Health, Aging and Body
Composition (Health ABC) study. j Am Geriatr Soc 2005,
53:1675-1680.
35. Ensrud KE, Ewing SK, Taylor BC, Fink HA, Cawthon PM, Stone KL, et
al.: Comparison of 2 frailty indexes for prediction of falls, dis-
ability, fractures, and death in older women. Arch Intern Med
2008, 168:382-389.
36. Fried LP, Tangen CM, Walston J, Newman AB, Hirsch C, Gottdiener
J, et al.: Frailty in older adults: evidence for a phenotype. j Ger-
ontol A Biol Sci Med Sci 2001, 56:M 146-M 156.
37. Ostir GV, Markides KS, Black SA, Goodwin JS: Lower body func-
tioning as a predictor of subsequent disability among older


Mexican Americans. J Gerontol A Biol Sci Med Sci 1998,
53:M491-M495.
38. Rolland Y, Lauwers-Cances V, Cesari M, Vellas B, Pahor M, Grandjean
H: Physical performance measures as predictors of mortality
in a cohort of community-dwelling older French women. Eur
J Epidemiol 2006, 21:1 13-122.
39. Stenback A: Physical health and physical disease as objective
fact and subjective experience. Arch Gen Psychiatry 1964,
I 1:290-301.
40. Markides KS, Black SA, Ostir GV, Angel RJ, GuralnikJM, Lichtenstein
M: Lower body function and mortality in Mexican American
elderly people. j Gerontol A Biol Sci Med Sci 2001,
56(4):M243-M247.
41. Rantanen T, Volpato S, Ferrucci L, Heikkinen E, Fried LP, GuralnikJM:
Handgrip strength and cause-specific and total mortality in
older disabled women: exploring the mechanism. j Am Geriatr
Soc 2003, 51:636-641.

Pre-publication history
The pre-publication history for this paper can be accessed
here:


http://www.biomedcentral.com/1471-2318/8/34/prepub


Page 9 of 9
(page number not for citation purposes)


BMC Geriatrics 2008, 8:34


Publish with BioMed Central and every
scientist can read your work free of charge
"BioMed Central will be the most significant development for
disseminating the results of biomedical research in our lifetime."
Sir Paul Nurse, Cancer Research UK
Your research papers will be:
available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours you keep the copyright
Submit your manuscript here: BioMedcentral
http://www.biomedcentral.com/info/publishing adv.asp




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs