Consequences of chronic obstructive pulmonary disease and chronic heart failure: the relationship between objective and subjective health.

Abstract

This study investigates whether the relationship between objective health parameters and general health perceptions was mediated by symptoms of dyspnoea and physical functioning in patients with chronic obstructive pulmonary disease (COPD) and patients with chronic heart failure (CHF). The different health parameters were organised according to Wilson and Cleary's conceptual model of patient outcomes (Wilson & Cleary (1995). Journal of the American Medical Association, 273, 59-65). Second, we investigated whether perceptions of personal control were related to the health parameters in the model. Consecutive patients with COPD and CHF were included from the outpatient clinics of a university hospital and a general hospital, and from a rehabilitation centre, all in the Netherlands. Ninety-five COPD patients (aged 65.0+/-9.3; forced expiratory volume in 1s (FEV1)<70%) were included and compared with 90 CHF patients (aged 59.6+/-10.0; left ventricular ejection fraction (LVEF)<45%). The relationship between objective health parameters (FEV1 or LVEF) and subjective health (self-reported physical functioning) was not mediated by symptoms of dyspnoea. FEV1 or LVEF and symptoms of dyspnoea were independently related to self-reported physical functioning, which was directly related to general health perceptions. Perceived health competence was related to symptoms of dyspnoea and general health perceptions in patients with either COPD or CHF. Although patients with COPD reported lower levels in all self-reported health parameters in the model than the patients with CHF, this study showed that the relations between the health parameters in the model were comparable for COPD and CHF patients.

Full text

Page 1

Social Science & Medicine 61 (2005) 2144–2154
Consequences of chronic obstructive pulmonary disease and
chronic heart failure: The relationship between objective and
subjective health
Rosemarie Arnolda,b,?, Adelita V. Ranchora,b, Gerard H. Koe ¨ terc,
Mike J.L. de Jongsted, Robbert Sandermana,b
aNorthern Centre for Healthcare Research (NCH), University of Groningen Medical Center, Antonius Deusinglaan 1, P.O. Box 196,
9700 AD Groningen, The Netherlands
bDepartment of Health Sciences, University of Groningen Medical Center, The Netherlands
cDepartment of Pulmonary Medicine, University of Groningen Medical Center, The Netherlands
dDepartment of Cardiology, Thorax Center, University of Groningen Medical Center, The Netherlands
Available online 31 May 2005
Abstract
This study investigates whether the relationship between objective health parameters and general health perceptions
was mediated by symptoms of dyspnoea and physical functioning in patients with chronic obstructive pulmonary
disease (COPD) and patients with chronic heart failure (CHF). The different health parameters were organised
according to Wilson and Cleary’s conceptual model of patient outcomes (Wilson & Cleary (1995). Journal of the
American Medical Association, 273, 59–65). Second, we investigated whether perceptions of personal control were
related to the health parameters in the model. Consecutive patients with COPD and CHF were included from the
outpatient clinics of a university hospital and a general hospital, and from a rehabilitation centre, all in the Netherlands.
Ninety-five COPD patients (aged 65.079.3; forced expiratory volume in 1s (FEV1)o70%) were included and
compared with 90 CHF patients (aged 59.6710.0; left ventricular ejection fraction (LVEF)o45%). The relationship
between objective health parameters (FEV1or LVEF) and subjective health (self-reported physical functioning) was not
mediated by symptoms of dyspnoea. FEV1or LVEF and symptoms of dyspnoea were independently related to self-
reported physical functioning, which was directly related to general health perceptions. Perceived health competence
was related to symptoms of dyspnoea and general health perceptions in patients with either COPD or CHF. Although
patients with COPD reported lower levels in all self-reported health parameters in the model than the patients with
CHF, this study showed that the relations between the health parameters in the model were comparable for COPD and
CHF patients.
r 2005 Elsevier Ltd. All rights reserved.
Keywords: Chronic obstructive pulmonary disease; Congestive heart failure; Dyspnoea; Health status; Quality of life; The Netherlands
Introduction
The development of a chronic illness may affect
several aspects of a patient’s health (Global Initiative for
Chronic Obstructive Lung Disease, 2003; Hunt et al.,
ARTICLE IN PRESS
www.elsevier.com/locate/socscimed
0277-9536/$-see front matter r 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.socscimed.2005.04.025
?Corresponding author. Northern Centre for Healthcare
Research (NCH), University of Groningen Medical Center,
Antonius Deusinglaan 1, P.O. Box 196, 9700 AD Groningen,
The Netherlands. Tel.: +31503611990; fax: +31503632406.
E-mail address: r.arnold@chir.umcg.nl (R. Arnold).

Page 2

2001a). These consequences of illness can be regarded as
a process of illness progression, which normally starts at
the development of physiological or biological abnorm-
alities, resulting in symptoms and physical limitations
that are noticed and reported by the patients. Even-
tually, patients will have to face their inability to take
part in their usual activities, which will influence their
perceptions of their health and ultimately their general
well-being. This chain of consequences of chronic illness
runs from the objective effects of the illness (physiolo-
gical abnormalities), to health status (physical function-
ing), and ultimately to the subjective perception of
health (health perceptions and well-being). This study
investigated the relationships between objective and
subjective health in Chronic Obstructive Pulmonary
Disease (COPD) and Chronic systolic Heart Failure
(CHF) patients.
To study the impact of chronic illness, most clinical
trials originally used the biomedical model, which
assessed the effects of etiological factors on the
pathological process and eventually on the clinical
outcomes (Stucki & Sigl, 2003; Johnston & Pollard,
2001). The aim of these studies was to enhance diagnosis
and treatment of patients and was usually focused on
largely objective health indicators. Studies from the
social sciences, on the other hand, usually focus on the
consequences of disease as perceived by the patients
(Wood-Dauphinee, 2001; Wilson & Cleary, 1995), and
are mainly interested in the subjective consequences of
disease. Wilson and Cleary (1995) have presented a
conceptual model of patient outcomes, which integrates
the biomedical model with the social scientific model of
health. They organise different measures of health from
biological factors at one end of the continuum to
perceived health at the other end of the continuum.
Moreover, this model assumes specific causal relation-
ships between the various health concepts, without
reciprocal effects (Sullivan, Kempen, Van Sonderen, &
Ormel, 2000). The five levels in the model, representing
different health concepts, are (1) biological and physio-
logical variables, (2) symptom status, (3) functional
status, (4) general health perceptions, and (5) overall
quality of life (QoL).
The present study investigates the relationship be-
tween objective and subjective health in patients with
COPD and CHF. Both COPD and CHF are seriously
debilitating conditions that have a profound impact on
patients’ functional status (Gosker et al., 2003) and QoL
(Ferrer et al., 1997; Juenger et al., 2002). COPD is
characterised by airflow limitation resulting in a reduced
ventilatory capacity, and for most patients in shortness
of breath or dyspnoea (Boueri, Bucher-Bartelson,
Glenn, & Make, 2001). In many COPD patients, these
symptoms of dyspnoea lead to limitations in daily
activities. CHF may result from various cardiac
disorders and is manifested in symptoms such as
dyspnoea and fatigue (Berry & McMurray, 1999). The
consequences of CHF include a reduced exercise
capacity and fluid retention, which may lead to
pulmonary and peripheral oedema (Hunt et al., 2001b).
We used Wilson and Cleary’s model to organise the
health variables from objective to subjective health. The
biological or physiological variables from Wilson and
Cleary’s model were operationalised as pulmonary
function for COPD patients, indicated by the forced
expiratory volume in 1s (FEV1), an objective measure of
airflow limitation (Global Initiative for Chronic Ob-
structive Lung Disease, 2003). For patients with CHF,
objective physical functioning was operationalised as left
ventricular function, indicated by the left ventricular
ejection fraction (LVEF) (Hunt et al., 2001a). Further-
more, in this study, symptom status in Wilson and
Cleary’s model was operationalised as dyspnoea, since
symptoms of dyspnoea are the main symptoms experi-
enced by both COPD and CHF patients.
Our study was particularly aimed at two research
questions. First, we addressed the question of whether
the relationships between physiological parameters
(FEV1 for COPD and LVEF for CHF) and general
health perceptions were mediated by symptoms of
dyspnoea and physical functioning. Wilson and Cleary’s
model does not mention whether only indirect relation-
ships exist (for instance, physiological variables are
related to physical functioning through symptom status)
or whether direct relationships between the variables
(for example, between physiological variables and
physical functioning) exist as well. Only a few previous
studies have empirically tested (parts of) Wilson and
Cleary’s model (Sullivan et al., 2000; Cosby, Holzemer,
Henry, & Portillo, 2000; Janz et al., 2001), and only one
of them has studied the sequence from objective to
subjective health factors by means of structural equation
modelling (Sullivan et al., 2000). This study, by Sullivan
et al. supported Wilson and Cleary’s model to the extent
that the relationship between biological variables and
general health perceptions was found to be mediated by
symptoms and physical functioning. Apart from this
indirect effect of biological variables on general health
perceptions, Sullivan also reported direct relationships
between non-adjacent variables in the model, that is,
between biological variables and physical functioning,
between symptoms and general health perceptions, and
between biological variables and general health percep-
tions (Sullivan et al., 2000). Consequently, we hypothe-
sised the existence of direct as well as indirect
relationships between the health parameters. Therefore,
we adapted Wilson and Cleary’s model in order to
incorporate the direct relationships between the non-
adjacent variables as well (see Fig. 1) and investigated
whether this new model was applicable for both COPD
and CHF. With respect to clinical characteristics, COPD
and CHF are comparable to a certain extent. Both are
ARTICLE IN PRESS
R. Arnold et al. / Social Science & Medicine 61 (2005) 2144–2154
2145

Page 3

progressive illnesses (Spencer, Calverley, Sherwood-
Burge, & Jones, 2001; Ekman, Fagerberg, & Lundman,
2002) characterised by dyspnoea and a reduced exercise
capacity (Hamilton, Killian, Summers, & Jones, 1995).
Although the underlying causes of these disorders are
quite different and may have differential effects on
health and QoL, the processes through which disease
affects health for COPD and CHF patients may be
comparable. Therefore, we hypothesised that compar-
able relationships might be discovered between the
factors in the model for COPD and CHF patients.
The model we studied did not include the variable
overall QoL, since our aim was to study the relation-
ships between objective and subjective health. The four
variables in Wilson and Cleary’s model are all related to
health. Overall QoL, on the other hand, concerns QoL
in general and may also include factors not related to
health, since previous studies have shown that overall
QoL was only to a limited extent explained by health
factors (Sullivan et al., 2000; Arnold et al., 2004).
Previous research has revealed that the more sub-
jective health factors in Wilson and Cleary’s model, such
as general health perceptions, were not completely
explained by the more objective health factors in the
model (Sullivan et al., 2000). It is most likely that factors
other than health also contribute to patients’ percep-
tions of health. Wilson and Cleary solved this issue by
adding external factors to their model, indicated as
‘characteristics of the environment’ and ‘characteristics
of the individual’, which may affect the health concepts
at several levels of the model (Wilson & Cleary, 1995).
An important psychological variable possibly related to
several aspects of health is personal control. Personal
control refers to individuals’ beliefs about their ability to
exert control over their own lives (Gurin, Gurin &
Morrison, 1978). Previous research has shown that
during the development of a chronic illness a patient’s
perceptions of personal control are affected (Helgeson,
1992). Furthermore, a greater amount of personal
control was found to be related to improved well-being
(Skinner, 1996) and better physical functioning (Yates &
Belknap, 1991). Consequently, the second aim of this
study was to investigate at which level in the model
perceptions of personal control were related to the
health parameters for patients with COPD and CHF.
Although various forms of personal control have been
described, this study focused on control as it manifests
in perceived competence (Smith, Wallston, & Smith,
1995), which refers to the extent to which an individual
feels capable of effectively managing his or her health
outcomes. Since relationships have been found between
perceived health competence and more preventive health
behaviour like exercise, abstinence from smoking, and
weight maintenance (Gebhardt, van der Doef, & Paul,
2001), and improved physical functioning (Arora et al.,
2002; Bonetti et al., 2001), perceived health competence
is likely to be related to health at different levels in
Wilson and Cleary’s model.
Methods
Patients and study design
Consecutive patients with COPD or CHF were
recruited from a university hospital, a general hospital,
and a university rehabilitation centre from January 2001
to June 2002. Data from patients in the rehabilitation
centre were combined with data from patients in the
outpatient clinics for both COPD and CHF. The data
used in this study are a part of a larger study that studied
the effects of a rehabilitation programme and therefore
assessed the COPD and CHF patients both before and
after rehabilitation programme participation. The as-
sessment carried out after the patients had finished their
rehabilitation programme was used in this study since at
that moment the patients were better able to be
compared to the patients in the outpatient clinic than
before they attended the rehabilitation programme. In
this assessment, no significant differences were found
between patients from the rehabilitation centre and
patients from the outpatient clinics with respect to
symptoms of dyspnoea, self-reported physical function-
ing, and general health perceptions. Furthermore,
correlation coefficients between these variables were
comparable for patients in rehabilitation and patients
ARTICLE IN PRESS
Physiological variables
(FEV1 vs. LVEF)
Symptoms of
dyspnoea
Physical functioning
General health
perceptions
Fig. 1. Theoretical model, based on Wilson and Cleary (1995).
R. Arnold et al. / Social Science & Medicine 61 (2005) 2144–2154
2146

Page 4

from the outpatient clinics for COPD and CHF. Power
analyses showed that in order to achieve a power of .80
to detect a difference of 15 points (SD 25) on subscale
physical functioning of the Rand 36-item health survey as
being statistically significant, a sample size of 45 was
needed for each group. Since patients with COPD and
CHF were recruited in rehabilitation as well as in an
outpatient clinic setting, in both patient groups 45
patients were included in both settings.
Eligible patients received written information about
the study and an informed consent form. COPD patients
were included if they were (1) diagnosed with COPD, (2)
registered with a FEV1of less than 70% of the predicted
value, and (3) free from other pulmonary disease.
Patients with CHF were included if they were (1)
diagnosed with CHF, and (2) registered with a LVEF
less than 45%. Furthermore, patients with CHF were
excluded if they suffered from angina pectoris or atrium
fibrillation, if they had had heart transplant surgery, or
if they had recently (in the past year) had a myocardial
infarction. Patients in both groups were aged between 40
and 80, free from any other serious disease (such as
cancer), free from psychiatric problems over the
previous year, and fluent in the Dutch language. The
study had a cross-sectional design and data were
collected by means of self-report mailed questionnaires.
The Medical Ethics Committees of the centres approved
the study.
Assessments
Clinical variables: In the COPD group data on lung
function parameters (FEV1in litres, FEV1%pred and
FVC in litres) were collected from the patients’ medical
records. For CHF, information on LVEF was collected
from medical files. Furthermore, the percentages of
smokers in both groups and the number of pack-years
(number of years patients have smoked?number of
cigarettes a day/20) and illness duration were assessed.
Symptoms of dyspnoea: A short questionnaire was
used to measure the number of illness-related symptoms
experienced by the patients (Van der Molen et al., 2003).
This questionnaire, originally designed for COPD
patients, was adapted for patients with CHF based on
the literature (Hunt et al., 2001b) and in consultation
with cardiologists to incorporate items assessing mani-
festations of CHF. In this study, two items were selected
from the larger questionnaire assessing symptoms of
COPD or CHF. These items were: (1) How often in the
past week did you experience breathlessness at rest? (2)
How often in the past week did you experience breath-
lessness during exercise? These two items, ranging from
0 to 6, were averaged; higher scores indicate that
symptoms were more often experienced.
Functional status: The physical functioning subscale
(10 items) of the Rand 36-item Health Survey (Ware &
Sherbourne, 1992; Van der Zee, Sanderman, Heyink, &
De Haes, 1996) was selected for the assessment
of functional status. Physical functioning measures
the extent to which health interferes with daily acti-
vities such as climbing stairs. The scores on this subscale
vary from 0 to 100; higher scores indicate better
functioning.
General health perceptions: The general health percep-
tions subscale (5 items) of the Rand 36-item Health
Survey (Ware & Sherbourne, 1992; Van der Zee et al.,
1996) measures the perceptions people have concerning
their health in general. Again, scores on this subscale
vary from 0 to 100 and higher scores indicate better-
perceived health.
Perceived health competence: The Perceived Health
Competence Scale (Smith et al., 1995) measures the
degree to which an individual feels capable of effectively
controlling and managing his or her health outcomes.
This scale consists of four positively formulated items
and four negatively formulated items. All items add up
to a total score (range 8–40) with higher scores
indicating greater levels of competence. Examples of
items include ‘I handle myself well with respect to my
health’ and ‘No matter how hard I try, my health just
doesn’t turn out the way I would like’.
Socio-demographic variables: Age, gender, marital
status and educational level were assessed. Educational
level was assessed according to the International
Standard Classification of Education (United Nations
Educational Scientific and Cultural Organization, 1976;
Ranchor, Bouma, & Sanderman, 1996), which has six
categories that were recoded into four categories,
ranging from lower to higher educated.
Data analysis
The first part of the study concerned confirmative
analyses in order to study the relationship between
objective and subjective health factors. On the basis of
previous research, we adapted Wilson and Cleary’s
model to incorporate the direct and indirect relation-
ships between the variables, and investigated whether
this new model was applicable for patients with COPD
as well as patients with CHF.
First, t-tests andw2tests were performed to compare
the COPD and CHF patients with respect to demo-
graphic and clinical characteristics. Second, the differ-
ences between the groups in symptoms of dyspnoea,
physical functioning, general health perceptions and
perceived health competence were tested by means of
analysis of covariance, to control for age differences
between the groups. Third, bivariate correlation ana-
lyses were performed to study the associations of
dyspnoea, physical functioning, general health percep-
tions and perceived health competence in both patient
groups.
ARTICLE IN PRESS
R. Arnold et al. / Social Science & Medicine 61 (2005) 2144–2154
2147

Page 5

Fourth, the theoretical model was studied (see also
Fig. 1) by means of structural equation modelling.
Structural equation modelling is a multivariate technique
for investigating the tenability of a hypothesised model
and is an elaboration of regression analysis. In order to
be able to perform LISREL, two criteria must be met:
first, a causal relationship between two variables is
known or assumed on the basis of previous research or
theoretical considerations. Second, no reciprocal rela-
tionships or ‘feedback loops’ can be tested. LISREL 8.20
was used to study the hypothesised model. The goodness
of fit of the models was judged by three measures. The w2
test compares the model to the hypothesised model and
should therefore not be significant (p4:05). The root
mean square error of approximation (RMSEA) is an
index of model fit, indicated by a value below .05. The
expected cross validation index (ECVI) is an index taking
fit and model parsimony into account, thus prohibiting
the addition of too many irrelevant parameters to the
model. ECVI values should be low; values lower than the
ECVI for the saturated model in particular indicate a
good fit for the model (Browne & Cudeck, 1989).
The input for structural equation modelling was the
correlation matrix (Table 3) with the relationships
between FEV1or LVEF, dyspnoea, self-reported physi-
cal functioning, and general health perceptions. Separate
analyses were performed for patients with COPD and
CHF in which the same model was studied (illustrated in
Fig. 1). The non-significant paths between the variables
were removed one by one to obtain a model with optimal
fit and greatest parsimony. Models for both groups were
evaluated by examining the parameter estimates and
measures of overall fit provided by LISREL.
The second part of our study had a more exploratory
character and was aimed at investigating at which level in
Wilson and Cleary’s model perceived health competence
was related to the health factors in the model. Three
different regression analyses were performed for each
group. First, the extent to which symptoms of dyspnoea
could be explained by lung function or ejection fraction
and perceived health competence was assessed (after
correction for age). In the second analysis, this analysis
was repeated with self-reported physical functioning as
the dependent variable and in addition to dyspnoea as an
independent variable. Third, the same analysis was
performed with general health perceptions as the
dependent variable and in addition to self-reported
physical functioning as an independent variable.
Results
Patient groups
One hundred and thirty-four COPD patients (86 from
the outpatient clinic and 48 in rehabilitation) were asked
to participate in the study. In the outpatient clinics, of
the 86 potential participants, 19 did not respond and 11
refused to participate in the study. In the rehabilitation
centre, of the 48 potential participants, one patient died,
one had an exacerbation of COPD, and seven withdrew
from the study before the end of the rehabilitation
programme. Patients who were not included in the study
did not differ significantly from the participants with
respect to age and gender. In total, 95 COPD patients
were included in the study (response rate, 71%).
In the CHF patient group, 138 patients (110 from the
outpatient clinic and 28 in rehabilitation) were invited to
the study. Of the 110 potential participants in the
outpatient clinic, 45 refused to participate. In the
rehabilitation centre, of the 28 potential participants,
three withdrew from the study before the end of the
rehabilitation programme. Patients who were not
included did not differ significantly from the participants
with respect to gender, but a significant difference was
found with respect to age: patients refusing to partici-
pate were on average 8 years older than the participants.
Ninety CHF patients participated in the study (response
rate, 65%).
Socio-demographic and clinical variables
Table 1 shows that the COPD patients were on
average 5 years older (po:001) than the CHF patients
(65.0 and 59.6 years old, respectively). COPD patients
reported a significantly higher number of pack-years
(35.3) than the CHF patients (27.4). Furthermore, illness
duration for COPD patients was on average 3 years
longer than for CHF patients (10.2 against 6.5). Other
demographic variables did not differ significantly
between the groups.
Differences between COPD and CHF in self-reported
health and perceived health competence
Table 2 shows that the COPD patients scored
significantly worse than the CHF patients with respect
to dyspnoea, self-reported physical functioning, general
health perceptions, and perceived health competence,
after correction for age differences.
Relationships between clinical variables, dyspnoea,
physical functioning, and general health perceptions
The input for the structural equation modelling
consisted of the correlation matrix as presented in
Table 3 (with the exception of perceived health
competence). In both COPD and CFH patients, FEV1
or LVEF were only significantly related to self-reported
physical functioning (r ¼ :22 for both COPD and CHF).
Symptoms of dyspnoea, physical functioning and
ARTICLE IN PRESS
R. Arnold et al. / Social Science & Medicine 61 (2005) 2144–2154
2148

Page 6

general health perceptions were significantly interre-
lated.
Figs. 2 and 3 show the best fitting models for the
patients with COPD and CHF. The parameters for
goodness of fit indicate a good fit for the models for
both COPD and CHF, according to the w2values
(w2¼ .41, p ¼ :81, df ¼ 2 for COPD; w2¼ .71, p ¼ :87,
df ¼ 3 for CHF). In both patient groups, the RMSEA
value is lower than .05 (RMSEA ¼ .00 for COPD;
RMSEA ¼ .00
Furthermore, the ECVI values are lower than the ECVI
for the saturated model for the COPD and CHF patients
forCHF), indicatinga good fit.
in this study (ECVI ¼ .18 (sat. ¼ .22) for COPD;
ECVI ¼ .17 (sat. ¼ .23) for CHF), which also indicates
a good fit for both models.
FEV1 for COPD patients and LVEF for CHF
patients were not significantly related to the level of
dyspnoea. Nevertheless, FEV1 and LVEF showed a
significant relationship with self-reported physical func-
tioning (b ¼ :20 for COPD; b ¼ :17 for CHF). Further-
more, dyspnoea was significantly related to self-reported
physical functioning (b ¼ ?:63 for COPD; b ¼ ?:67 for
CHF) in both groups, whereas physical functioning was
significantly related to general health perceptions
ARTICLE IN PRESS
Table 2
Self-reported health and perceived health competence of patients with COPD and CHF
COPDCHF Difference between groups
Meana
SD Meana
SD
F-values
Dyspnoea
Physical functioning
General health perceptions
Perceived health competence
2.7
38.3
38.4
24.8
1.1
24.7
18.7
4.5
1.8
57.8
49.8
27.6
1.1
26.1
19.9
4.3
28.8*
19.9*
21.7*
14.2*
Analysis of variance between groups: *po.001 (adjusted for age differences).
aUnadjusted means (uncorrected for age differences between groups).
Table 1
Demographic variables and clinical characteristics of patients with COPD and CHF
COPD, N ¼ 95CHF, N ¼ 90Difference between groups
MeanSD Mean SD
p-value
Age (years)
Gender (% female)
Educational level (%)
Primary
Lower vocational
Secondary/intermediate vocational
Higher vocational/university
65.0
35.8
9.3 59.6
24.4
10.0.001a
n.s.
n.s.
39.0
32.6
15.8
12.6
26.7
35.6
23.3
14.4
Marital status (% with partner)
FEV1(l)
FEV1%pred
FVC (l)
LVEF (%)
Number of smokers (%)
Smokers
Former smokers
No smokers (never)
86.2
1.39
45.7
3.3
81.8n.s.
.7
19.0
.9
30.39.4
n.s.
17.9
74.7
7.4
14.4
74.4
11.2
Pack-years (years)
Illness duration (years)
35.3
10.2
22.2
10.0
27.4
6.5
22.1
7.9
.03b
.01c
n.s. ¼ not significant.
aDifferences between groups: t value ¼ 3.8.
bt value ¼ 2.2.
ct value ¼ 2.5.
R. Arnold et al. / Social Science & Medicine 61 (2005) 2144–2154
2149

Page 7

(b ¼ :39 for COPD; b ¼ :52 for CHF). Only in COPD
patients dyspnoea was directly related to general health
perceptions (b ¼ ?:37). In patients with COPD, general
health perceptions were explained by symptoms of
dyspnoea and self-reported physical functioning. For
patients with CHF, on the other hand, general health
perceptions were only explained by self-reported physi-
cal functioning.
The structural models were nearly identical for COPD
and CHF patients. In both patient groups, clinical
variables and symptoms of dyspnoea were directly
related to self-reported physical functioning, which in
turn was related to general health perceptions. The only
difference between both models is that in the COPD
patients, dyspnoea was directly related to general health
perceptions.
Relationships between perceived health competence and
the variables of health
The regression analyses assessed the extent to which
each health parameter in the model was explained by the
preceding parameters in the model and by perceived
health competence (see Table 4).
ARTICLE IN PRESS
Table 3
Pearson’s correlation coefficients of illness severity (FEV1vs. LVEF), dyspnoea, physical functioning (Rand-36), general health
perceptions (Rand-36), and perceived health competence for patients with COPD and CHF
DyspnoeaPhysical functioningGeneral health perceptions Perceived health competence
Illness severity
FEV1
LVEF
COPD
CHF
?.04
?.08
.22*
.22*
.14
.15
.10
.06
Dyspnoea COPD
CHF
?.63***
?.69***
?.62***
?.37***
.63***
.52***
?.41***
?.32**
.29**
.40***
Physical functioning COPD
CHF
General health perceptionsCOPD
CHF
.39***
.49***
?po:05;??po:01;???po:001.
FEV1
Dyspnoea Physical functioning
General health
perceptions
0.20
-0.37
-0.63
0.39
Fig. 2. Final model of relationships between FEV1, dyspnoea, physical functioning and general health perceptions for patients with
COPD. w2¼ .41 (p ¼ :81; df ¼ 2); RMSEA ¼ .00; ECVI ¼ .18 (sat. ¼ .22); N ¼ 93.
LVEF
Dyspnoea Physical functioning
General health
perceptions
0.17
-0.67
0.52
Fig. 3. Final model of relationships between LVEF, dyspnoea, physical functioning and general health perceptions for patients with
CHF. w2¼ .71 (p ¼ :87; df ¼ 3); RMSEA ¼ .00; ECVI ¼ .17 (sat. ¼ .23); N ¼ 88.
R. Arnold et al. / Social Science & Medicine 61 (2005) 2144–2154
2150

Page 8

The results of the regression analyses showed that
perceived health competence contributed to the explana-
tion of dyspnoea, although the variables age and FEV1
or LVEF did not contribute to dyspnoea. Only in CHF
patients did perceived health competence contribute to
the explanation of physical functioning after the other
variables had been entered into the equation. Finally, in
both patient groups, perceived health competence
contributed significantly to general health perceptions
after the variables age, FEV1or LVEF, dyspnoea, and
physical functioning had been entered into the equation.
Therefore, perceived health competence contributed
significantly to the explanation of dyspnoea and general
health perceptions for both COPD and CHF patients,
although perceived health competence was only related
to physical functioning in CHF patients.
Discussion
This study investigated the relationship between
objective and subjective health parameters in patients
with COPD and CHF. The results support our
hypothesis that direct as well as indirect relationships
between the health parameters exist (see Fig. 1),
although not all hypothesised relationships between
the variables in the model were found in the groups of
patients with COPD and CHF. Consistent with the
hypothesised model, the relationship of FEV1or LVEF
and symptoms of dyspnoea to general health percep-
tions was mediated by self-reported physical functioning
for COPD and CHF patients. However, the relationship
between FEV1 or LVEF and self-reported physical
functioning was not mediated by symptoms of dys-
pnoea, although FEV1 or LVEF and symptoms of
dyspnoea were independently related to self-reported
physical functioning. Furthermore, no direct relation-
ships were found between FEV1and LVEF and general
health perceptions, and dyspnoea was only directly
related to general health perceptions for patients with
COPD. Consistent with our hypothesis was the result
that the structural models were nearly identical for
COPD and CHF patients.
In contrast to Wilson and Cleary’s model (1995), our
results did not reveal a relationship between physiolo-
gical parameters and symptom status. The results we
found, though, were consistent with previous studies,
which also reported weak correlations between pulmon-
ary or left ventricular function and symptoms of
dyspnoea (Wijnhoven, Kriegsman, Hesselink, Penninx,
& de Haan, 2001; Singh, Sodergren, Hyland, Williams,
& Morgan, 2001; Herlitz et al., 2000). Furthermore,
Wilson and Cleary themselves argued that in some
disorders, biological or physiological parameters may be
unrelated to symptom status. These results indicate that
pulmonary or left ventricular function and symptoms of
dyspnoea are two different, weakly related aspects of the
disorders, although both are related to the physical
limitations that patients experience.
The second aim of this study was to investigate at
which level in the model perceived health competence
was related to health factors in Wilson and Cleary’s
ARTICLE IN PRESS
Table 4
Regression analysis of general health perceptions (Rand-36) on perceived health competence, illness severity (FEV1vs. LVEF),
dyspnoea, and physical functioning for patients with COPD (n ¼ 95) and patients with CHF (n ¼ 90)
Predictors DyspnoeaPhysical functioning General health perceptions
B
SE B
Beta
B
SE B
Beta
B
SE B
Beta
COPD
Age
FEV1
Dyspnoea
Physical functioning
Perceived health competence
?.01
.00
—
—
?.11
.01
.01
?.12
.02
—
—
?.41***
.18
?.72
.31
?14.32
—
?.05
.20
.10
1.79
—
.46
?.27***
.24**
?.66***
—
?.01
.51
.51 .16
.08
1.74
.08
.34
.25**
?.03
?.21*
.49***
.17*
?.03
?3.49—
— .37
.71 .03
Cum. R2
.54
CHF
Age
LVEF
Dyspnoea
Physical functioning
Perceived health competence
.01.01
.01
.12
?.17
.47
?13.87
—
1.23
.20
.21
1.83
—
.50
?.07
.17*
?.60***
—
.19*
.65
.06
.16
.18
1.94
.09
.42
.33***
.03
?.02
.42***
.34***
?.01
—
—
?.08
?.06
—
—
?.30**
.12
—
—
?.39
.32
1.61.03
Cum. R2
.54.47
?po:05;??po:01;???po:001.
R. Arnold et al. / Social Science & Medicine 61 (2005) 2144–2154
2151

Page 9

model. In both COPD and CHF, perceived health
competence was related to dyspnoea and general health
perceptions once all other health concepts were entered
into the equation. Only in CHF patients was perceived
health competence significantly related to self-reported
physical functioning. Therefore, perceptions of personal
control were related to health factors at several levels in
the model for both COPD and CHF patients. Perceived
health competence in particular was related to symp-
toms of dyspnoea and general health perceptions in both
patient groups.
As the relationships between perceived health compe-
tence and several health parameters were cross-section-
ally studied, no firm conclusions can be drawn
concerning the causality of these relationships. Recipro-
cal relationships between health parameters and percep-
tions of competence are very likely to exist (Smith et al.,
1995). The relationships between symptoms of dyspnoea
and perceived health competence and between perceived
health competence and general health perceptions found
in this study suggest the possibility that perceived health
competence is an intervening variable between symp-
toms of dyspnoea and general health perceptions. It
seems plausible that patients who experience more
dyspnoea also have a lower level of perceived health
competence, which in turn affects their general health
perceptions.
A few remarks concerning the methods of this study
should be made. Structural equation modelling is a
technique to study the tenability of a hypothesised
model. Although structural equation modelling pro-
duces directional models, it is unable to test causal
relationships between the variables when applied to
cross-sectional data. Therefore, it cannot be determined
whether the different health variables in our model do
indeed follow each other chronologically. It would be of
interest to study the relationships between these
variables longitudinally and, ideally, to include patients
from the outset of the development of their illness.
Second, in this study, data from patients in the
outpatient clinics were combined with data from
patients who had just finished their rehabilitation
programme. Therefore, as mentioned in the methods
section, the possible influence on the results of using
data from two different groups was checked and turned
out not to be the case. The advantage of including
patients in rehabilitation and patients from the out-
patient clinics is that the patient samples in this study
may better represent the total populations of COPD and
CHF patients. Third, from the potential CHF patients
in the outpatient clinic 45 patients refused to participate
in the study. These patients were not willing to
participate, for which the particular reasons are not
known. The result that the CHF patients who were not
willing to participate were on average about 8 years
older than the participants, indicates that younger
patients were more inclined to participate than older
patients (selection bias). As a result, the COPD patients
were on average five years older than the CHF patients.
Therefore, the analyses were corrected for age differ-
ences between groups. Fourth, the present study did not
have a very large sample size, which may have caused
some limitations. In particular, the limited sample size of
this study did not allow cross-validation of the model on
independent data, which limits generalisation of the
study results.
This study has both theoretical and clinical implica-
tions. Of theoretical importance is the fact that Wilson
and Cleary’s model of health was empirically studied,
something done by only a handful of earlier studies
(Sullivan et al., 2000; Cosby et al., 2000; Janz et al.,
2001). Our study provides further insight into the
process through which objective and subjective health
parameters contribute to general health perceptions.
Moreover, this process was studied in two different
patient groups. Only a few recent studies have compared
different chronic disorders with respect to health
(Sprangers et al., 2000; Stavem, Lossius, Kvien, &
Guldvog, 2000; Kempen, Ormel, Brilman, & Relyveld,
1997; Arnold et al., 2004; Schlenk et al., 1998), yielding
inconsistent results. In this study, COPD patients
reported worse functioning than CHF patients with
respect to dyspnoea, physical functioning, and general
health perceptions. Irrespective of these differences, this
study showed that the relationships between the
different health factors in the model are largely similar
for COPD and CHF patients.
Relevant to clinical practice is the finding that general
health perceptions were explained by both self-reported
physical functioning and perceived health competence.
Apparently, not only physical functioning but also
perceptions of personal control are strongly related to
patients’ perceptions concerning their general health.
This indicates that paying explicit attention in the
treatment of these patients to the improvement of
perceived health competence may lead to improvements
in health parameters. Furthermore, objective health
parameters were not directly related to general health
perceptions. This indicates that it is not the physical
disorder per se but rather the resultant limitations in
physical functioning that are important determinants of
the general health perceptions of patients. Though the
assessment of pulmonary function or left ventricular
function provides an accurate indication of disease
severity (Global Initiative for Chronic Obstructive Lung
Disease, 2003; Hunt et al., 2001a), it may be more
relevant for a clinician to know the physical limitations
patients experience and their perceptions of health
competence as these factors are directly related to
patients’ health perceptions and their well-being (Wilson
& Cleary, 1995), and therefore may be directive for the
treatment of the patients.
ARTICLE IN PRESS
R. Arnold et al. / Social Science & Medicine 61 (2005) 2144–2154
2152

Page 10

To conclude, this study used Wilson and Cleary’s
model to assess the relationship between objective and
subjective health parameters, and to investigate the
relationship between perceived health competence with
the health parameters at different levels in their model.
Although COPD patients showed lower levels of
functioning than the CHF patients in all self-reported
health parameters in the model, this study showed that
the relationships between the health parameters in the
model were largely comparable between COPD and
CHF patients.
References
Arnold, R., Ranchor, A. V., Sanderman, R., Kempen, G. I. J.
M., Ormel, J., & Suurmeijer, T. P. B. M. (2004). The relative
contribution of domains of quality of life to overall quality
of life for different chronic diseases. Quality of Life
Research, 13, 883–896.
Arora, N., Johnson, P., Gustafson, D., McTavish, F., Hawkins,
R., & Pingree, S. (2002). Barriers to information access,
perceived health competence, and psychosocial health
outcomes: Test of a medication model in a breast cancer
sample. Patient Education and Counselling, 47, 37–46.
Berry, C., & McMurray, J. (1999). A review of quality-of-life
evaluations in patients with congestive heart failure.
Pharmacoeconomics, 16, 247–271.
Bonetti, D., Johnston, M., Rodriguez-Marin, J., Pastor, M.,
Martin-Aragon, M., Doherty, E., & Sheehan, K. (2001).
Dimensions of perceived control: A factor analysis of three
measures and an examination of their relation to activity
level and mood in a student and cross-cultural patient
sample. Psychology and Health, 16, 655–674.
Boueri, F. M. V., Bucher-Bartelson, B. L., Glenn, K. A., &
Make, B. J. (2001). Quality of life measured with a generic
instrument (Short Form-36) improves following pulmonary
rehabilitation in patients with COPD. Chest, 119, 77–84.
Browne, M. W., & Cudeck, R. (1989). Single sample cross-
validation indices for covariance structures. Multivariate
Behavioral Research, 24, 445–455.
Cosby, C., Holzemer, W. L., Henry, S. B., & Portillo, C. J.
(2000). Hematological complications and quality of life in
hospitalized AIDS patients. Aids Patient Care and STDs,
14, 269–279.
Ekman, I., Fagerberg, B., & Lundman, B. (2002). Health-
related quality of life and sense of coherence among elderly
patients with severe chronic heart failure in comparison with
healthy controls. Heart & Lung, 31, 94–101.
Ferrer, M., Alonso, J., Morera, J., Marrades, R. M., Khalaf,
A., Aguar, M. C., Plaza, V., Prieto, L., & Anto, J. M.
(1997). Chronic obstructive pulmonary disease stage and
health-related quality of life. The Quality of Life of Chronic
Obstructive Pulmonary Disease Study Group. Annals of
Internal Medicine, 127, 1072–1079.
Gebhardt, W. A., van der Doef, M. P., & Paul, L. B. (2001).
The Revised Health Hardiness Inventory (RRHI-24):
psychometric properties and relationship with self-reported
health and health behavior in two Dutch samples. Health
Education Research, 16, 579–592.
Global Initiative for Chronic Obstructive Lung Disease. (2003).
Global strategy for the diagnosis, management, and preven-
tion of Chronic Obstructive Pulmonary Disease. NHLBI/
WHO workshop report, Update of the management
sections, National Heart, Lung and Blood Institute, 2003
July 1 [cited 2004 March 4]. Available from: http://
www.goldcopd.com.
Gosker, H. R., Lencer, N. H. M. K., Franssen, F. M. E., Van
der Vusse, G. J., Wouters, E. F. M., & Schols, A. M. W.
(2003). Striking similarities in systemic factors contributing
to decreased exercise capacity in patients with severe chronic
heart failure or COPD. Chest, 123, 1416–1424.
Gurin, P., Gurin, G., & Morrison, B. M. (1978). Personal and
ideological aspects of internal and external control. Social
Psychology, 41, 275–296.
Hamilton, A. L., Killian, K. J., Summers, E., & Jones, N. L.
(1995). Muscle strength, symptom intensity, and exercise
capacity in patients with cardiorespiratory disorders. Amer-
ican Journal of Respiratory and Critical Care Medicine, 152,
2021–2031.
Helgeson, V. S. (1992). Moderators of the relation between
perceived control and adjustment to chronic illness. Journal
of Personality and Social Psychology, 63, 656–666.
Herlitz, J., Wiklund, I., Sjoland, H., Karlson, B. W., Karlsson,
T., Haglid, M., Hartford, M., & Caidahl, K. (2000). Relief
of symptoms and improvement of quality of life five years
after coronary artery bypass grafting in relation to
preoperative ejection fraction. Quality of Life Research, 9,
467–476.
Hunt, S. A., Baker, D. W., Chin, M. H., Cinquegrani, M. P.,
Feldman, A. M., Francis, G. S., Ganiats, T. G., Goldstein,
S., Gregoratos, G., Jessup, M. L., Noble, R. J., Packer, M.,
Silver, M. A., & Stevenson, L. W. (2001a). ACC/AHA
guidelines for the evaluation and management of chronic heart
failure in the adult: a report of the American College of
Cardiology/American Heart Association task force on
practice guidelines (committee to revise the 1995 guidelines
for the evaluation and management of heart failure).
American College of Cardiology Website [cited 2004 March
5]. Available from: http://www.acc.org/clinical/guidelines/
failure/hf_index.htm.
Hunt, S. A., Baker, D. W., Chin, M. H., Cinquegrani, M. P.,
Feldman, A. M., Francis, G. S., Ganiats, T. G., Goldstein,
S., Gregoratos, G., Jessup, M. L., Noble, R. J., Packer, M.,
Silver, M. A., Stevenson, L. W., Gibbons, R. J., Antman, E.
M., Alpert, J. S., Faxon, D. P., Fuster, V., Gregoratos, G.,
Jacobs, A. K., Hiratzka, L. F., Russell, R. O., & Smith, S.
C. J. (2001b). ACC/AHA guidelines for the evaluation and
management of chronic heart failure in the adult: Executive
summary. Circulation, 104, 2996–3007.
Janz, N. K., Janevic, M. R., Dodge, J. A., Fingerlin, T. E.,
Schork, M. A., Mosca, L. J., & Clark, N. M. (2001). Factors
influencing quality of life in older women with heart disease.
Medical Care, 39, 588–598.
Johnston, M., & Pollard, B. (2001). Consequences of disease:
testing the WHO international classification of impairments,
disabilities and handicaps (ICIDH) model. Social Science &
Medicine, 53, 1261–1273.
Juenger, J., Schellberg, D., Kraemer, S., Haunstetter, A.,
Zugck, C., Herzog, W., & Haass, M. (2002). Health related
quality of life in patients with congestive heart failure:
ARTICLE IN PRESS
R. Arnold et al. / Social Science & Medicine 61 (2005) 2144–2154
2153

Page 11

Comparison with other chronic diseases and relation to
functional variables. Heart, 87, 235–241.
Kempen, G. I. J. M., Ormel, J., Brilman, E. I., & Relyveld, J.
(1997). Adaptive responses among Dutch elderly: The
impact of eight chronic medical conditions on health-related
quality of life. American Journal of Public Health, 87, 38–44.
Ranchor, A. V., Bouma, J., & Sanderman, R. (1996).
Vulnerability and social class: Differential patterns of
personality and social support over the social class.
Personality and Individual Differences, 20, 229–237.
Schlenk, E. A., Erlen, J. A., Dunbar Jacob, J., McDowell, J.,
Engberg, S., Sereika, S. M., Rohay, J. M., & Bernier, M. J.
(1998). Health-related quality of life in chronic disorders: A
comparison across studies using the MOS SF-36. Quality of
Life Research, 7, 57–65.
Singh, S. J., Sodergren, S. C., Hyland, M. E., Williams, J., &
Morgan, M. D. (2001). A comparison of three disease-
specific and two generic health-status measures to evaluate
the outcome of pulmonary rehabilitation in COPD.
Respiratory Medicine, 95, 71–77.
Skinner, E. A. (1996). A guide to constructs of control. Journal
of Personality and Social Psychology, 71, 549–570.
Smith, M. S., Wallston, K., & Smith, C. (1995). The
development and validation of the Perceived Health
Competence Scale. Health Education Research, 10, 51–64.
Spencer, S., Calverley, P. M., Sherwood-Burge, P., & Jones, P.
W. (2001). Health status deterioration in patients with
chronic obstructive pulmonary disease. American Journal of
Respiratory and Critical Care Medicine, 163, 122–128.
Sprangers, A. G., De Regt, E. B., Andries, F., Van Agt, H. M.
E., Bijl, R. V., De Boer, J. B., Foets, M., Hoeymans, N.,
Jacobs, A. E., Kempen, G. I. J. M., Miedema, H. S., Tijhuis,
M. A. R., & De Haes, H. C. J. M. (2000). Which chronic
conditions are associated with better or poorer quality of
life? Journal of Clinical Epidemiology, 53, 895–907.
Stavem, K., Lossius, M. I., Kvien, T. K., & Guldvog, B. (2000).
The health-related quality of life of patients with epilepsy
compared with angina pectoris, rheumatoid arthritis,
asthma and chronic obstructive pulmonary disease. Quality
of Life Research, 9, 865–871.
Stucki, G., & Sigl, T. (2003). Assessment of the impact of
disease on the individual. Best Practice and Research in
Clinical Rheumatology, 17, 451–473.
Sullivan, M. D., Kempen, G. I. J. M., Van Sonderen, E., &
Ormel, J. (2000). Models of health-related quality of life in a
population of community-dwelling Dutch elderly. Quality of
Life Research, 9, 801–810.
United Nations Educational Scientific and Cultural Organiza-
tion. (1976). International Standard Classification of Educa-
tion (ISCED). Paris, France: UNESCO.
Van der Molen, T., Willemse, B. M. W., Schokker, S., Ten
Hacken, N. H. T., Postma, D., & Juniper, E. F. (2003).
Development, validity and responsiveness of the Clinical
COPD Questionnaire. Health and Quality of Life Outcomes
[serial online] 2003 April 28 [cited 2003 Aug 5], 1, [10
screens]. Available from: URL: http://www.hqlo.com/con-
tent/1/1/13.
Van der Zee, K., Sanderman, R., Heyink, J. W., & De Haes, H.
(1996). Psychometric qualities of the RAND 36-item Health
Survey 1.0: A multidimensional measure of general health
status. International Journal of Behavioral Medicine, 3,
104–122.
Ware, J. E., & Sherbourne, C. D. (1992). The MOS
36-item Short-Form Health Survey (SF-36): I. Conce-
ptual framework and item selection. Medical Care, 30,
473–483.
Wijnhoven, H. A., Kriegsman, D. M., Hesselink, A. E.,
Penninx, B. W., & de Haan, M. (2001). Determinants of
different dimensions of disease severity in asthma and
COPD: Pulmonary function and health-related quality of
life. Chest, 119, 1034–1042.
Wilson, I. B., & Cleary, P. D. (1995). Linking clinical variables
with health-related quality of life. A conceptual model of
patient outcomes. Journal of the American Medical Associa-
tion, 273, 59–65.
Wood-Dauphinee, S. L. (2001). Assessment of back-related
quality of life: The continuing challenge. Spine, 26, 857–861.
Yates, B. C., & Belknap, D. C. (1991). Predictors of physical
functioning after a cardiac event. Heart & Lung, 20,
383–390.
ARTICLE IN PRESS
R. Arnold et al. / Social Science & Medicine 61 (2005) 2144–2154
2154