Qual Life Res
DOI 10.1007/s11136-017-1503-y
Lower educational level and unemployment increase the impact
of cardiometabolic conditions on the quality of life: results
of a population-based study in South Australia
David Alejandro  González‑Chica1 · Robert Adams2 · Eleonora Dal Grande3 ·
Jodie Avery3 · Phillipa Hay4 · Nigel Stocks1 
Accepted: 10 January 2017
© Springer International Publishing Switzerland 2017
Abstract  (95%CI 44.4–45.9) and 39.1 (95%CI 37.7–40.5) among
Purpose  To investigate if sociodemographic characteris-
tics increase the adverse effects of cardiovascular diseases
(CVD) and cardiometabolic risk factors (CMRF) on health-
related quality of life (HRQoL).
Methods  Cross-sectional, face-to-face survey investigat-
ing 2379 adults living in South Australia in 2015 (57.1 ± 14
years; 51.7% females). Questions included diagnosis of
CMRF (obesity, diabetes, hypertension, dyslipidaemia) and
CVD. Physical and mental HRQoL were assessed using the
SF-12v1 questionnaire. Multiple linear regression models
including confounders (sociodemographic, lifestyle, use of
preventive medication) and interaction terms between soci-
odemographic variables and cardiometabolic conditions
were used in adjusted analysis.
Results  The prevalence of CMRF (one or more) was
54.6% and CVD was 13.0%. The physical HRQoL reduced
from 50.8 (95%CI 50.2–51.4) in healthy individuals to 45.1
Electronic supplementary material  The online version of this
article (doi:10.1007/s11136-017-1503-y) contains supplementary
material, which is available to authorized users.
* David Alejandro González‑Chica
david.gonzalez@adelaide.edu.au
Robert Adams
robert.adams@adelaide.edu.au
Eleonora Dal Grande
eleonora.dalgrande@adelaide.edu.au
Jodie Avery
jodie.avery@adelaide.edu.au
Phillipa Hay
P.Hay@westernsydney.edu.au
Nigel Stocks
nigel.stocks@adelaide.edu.au
those with CMRF and CVD, respectively. Adjustment for
sociodemographic variables reduced these differences in
33%, remaining stable after controlling for lifestyle and use
of preventive medications (p < 0.001). Differences in physi-
cal HRQoL according to cardiometabolic conditions were
twice as high among those with lower educational level,
or if they were not working. Among unemployed, having
a CMRF or a CVD had the same impact on the physical
HRQoL (9.7 lower score than healthy individuals). The
inverse association between cardiometabolic conditions and
mental HRQoL was subtle (p = 0.030), with no evidence of
disparities due to sociodemographic variables.
Conclusions  A lower educational level and unemploy-
ment increase the adverse effects of cardiometabolic con-
ditions on the physical HRQoL. Targeted interventions for
reducing CMRF and/or CVD in these groups are necessary
to improve HRQoL.
Keywords  Quality of life · Cardiovascular disease ·
Metabolic disease · Socioeconomic factors · Health status
disparities
1
Discipline of General Practice, School of Medicine, NHMRC
Centre of Research Excellence to Reduce Inequality in Heart
Disease, The University of Adelaide, 178 North Terrace,
Level 11, MDP DX 650 550, Adelaide, SA 5005, Australia
2
The Health Observatory, Discipline of Medicine, The
University of Adelaide, The Queen Elizabeth Hospital
Campus, Woodville, SA, Australia
3
Population Research and Outcome Studies, Discipline
of Medicine, School of Medicine, The University
of Adelaide, Adelaide, SA, Australia
4
Centre for Health Research, School of Medicine, University
of Western Sydney, Sydney, NSW, Australia
13
Vol.:(0123456789)

Introduction
Cardiovascular disease (CVD) is the leading cause of
global morbidity and mortality, responsible for the larg-
est fraction of the 36 million annual deaths related to non-
communicable chronic disease [1, 2]. In 2011, an estimated
3.7 million Australians (21%) adults had at least one CVD,
with 69,900 acute coronary events occurring in that year
[3]. Some cardiometabolic risk factors (CMRF) increase
the lifetime risk of experiencing cardiovascular events,
such as hypertension, diabetes mellitus, dyslipidaemia,
and obesity [4]. Worldwide, the prevalence of these CMRF
have increased over the last 30 years [2].
Although improved management of acute CVD was
responsible for a decline in mortality rates, disability-
adjusted life years (DALYs) due to CVD increased 23%
between 1990 and 2010 [1, 2, 5]. Additionally, low adher-
ence to the long-term management of CVD increased the
frequency of complications and health expenditure [6, 7].
Given this resource burden, other approaches are necessary
to improve clinical management, prevent complications,
and improve the quality of life (QoL) among individuals
with CVD [6–8].
QoL is a social construct, which represents an indi-
vidual’s perception of different domains of their life over
time, including physical, mental, social, environmental, and
general status that they experience [9]. Diverse studies have
demonstrated the impact of non-communicable chronic dis-
ease on health-related QoL (HRQoL), but despite their rel-
evance to CVD outcomes, they have not been investigated
in most studies [8]. HRQoL has become an increasingly
important patient-centred outcome in the study of CVD,
not only because its relationship with functional capacity
and adherence to treatment, but also for its role as a com-
plementary tool for monitoring disease severity and pro-
gression [8, 10, 11].
Furthermore, socioeconomic and demographic charac-
teristics (i.e. lower educational level, unemployment, and
female gender) have been associated with a lower HRQoL
[12–15]. Nevertheless, few studies have investigated these
variables as a source of HRQoL inequalities among peo-
ple with cardiometabolic conditions in representative pop-
ulation-based samples [14–16]. All of these studies used
a HRQoL instrument that provided a unique score, affect-
ing the capacity to identify the most impaired domain of
HRQoL. Therefore, the purpose of this paper was to inves-
tigate the extent to which sociodemographic characteristics
increase the adverse effects of cardiometabolic conditions
(CMRF and CVD) on the physical and mental domains of
HRQoL in a population-based sample of adults in South
Australia (SA). A second aim was to explore if the associa-
tions of these conditions with HRQoL are confounded by
lifestyle variables and the use of preventive medications.
13
Qual Life Res
Methods
We report a cross-sectional, face-to-face respondent-
based interview study (2015 Spring Health Omnibus
Survey) analysing data from a representative sample of
adults in SA [17]. This state is the fifth largest in Aus-
tralia (1.6  million inhabitants; 73% in the capital city),
with an average life expectancy of 83.0 years and a
human development index of 0.907, equivalent to that of
Sweden and the UK [18].
A multistage sampling process was used to select the
participants. In the first stage, cities/localities were strati-
fied according to their location (metropolitan or rural).
Rural localities with a population size of at least 10,000
and rural centres with 1,000 or more inhabitants were con-
sidered eligible. In the second stage, 530/3939 of the SA
Level 1 areas (SA1, smallest division unit used in the Aus-
tralian census in 2011) [18] were systematically selected
(132 in rural areas), with the probability of selection pro-
portional to their size. These methods allowed obtaining
self-weighting country samples. In the third stage, clusters
of 10 dwellings were systematically chosen in each selected
SA1. Finally, one person aged 15 years or over residing in
the household was randomly selected (the last individual to
have their birthday).
Of the 5,300 selected households, 58 were excluded
because they were vacant houses/businesses, 77 because
it was impossible to gain access to the building, and 748
because the contact was not established (six visits at dif-
ferent moments of day/evening and different days of the
week). Individuals were also excluded if terminally ill/men-
tal incapacity (n = 104) or when unable to speak English
(n = 87). Of the 4,226 eligible participants, 1,221 refused
to complete the interview, providing a final sample of 3,005
individuals (participation rate 71.1%).
Only individuals aged 35 years or over were included in
our study (n = 2379), considering the very low prevalence
of CVD among younger age groups. The power of the
study was estimated a posteriori considering the available
sample size. For a prevalence of CVD of 13%, fixing the
alpha in 5%, the power in 80%, and with a design effect of
1.3 it would be possible to detect a mean difference of at
least 2.2 points in the physical component of HRQoL (for
a standard deviation = 11.0) and 1.8 points in the mental
component (for a standard deviation = 9.1).
All interviews were administered face-to-face between
September and December 2015. Different quality control
methods were used, including a pilot study of 50 inter-
views to test the instrument, double data entry, and phone
verification of missing responses. Additionally, 10% of
participants were randomly selected, re-contacted, and
asked some key questions of the original survey for quality
control.
Qual Life Res
Sociodemographic variables collected in these question-
naires included gender, age, marital status, dwelling char-
acteristics, educational level, working status, and postcode
area of residence. An area-based composite score, the 2011
Australian Socio-Economic Indexes for Areas (SEIFA)
Index of Relative Socio-economic Advantage and Disad-
vantage (IRSAD), was used to provide a measure of socio-
economic position (SEP) at the postcode level. This index
is based on a range of census variables and is an indica-
tor of relative economic and social advantage/disadvantage
of people and households within an area [19]. High scores
indicate the respondent residing in a more advantaged area
and low scores indicate least advantage area.
Health variables were also collected, including lifestyle
variables (daily fruit and vegetable consumption, weekly
frequency of physical activity, alcohol intake, and smoking
status), self-reported weight and height, and previous diag-
nosis (“have you ever told by a doctor that you have…”)
and/or treatment (“are you on medication for…”) for dia-
betes, hypertension, dyslipidaemia, myocardial infarction,
angina, heart failure, atrial fibrillation/arrhythmia, and
stroke.
Outcome
The Medical Outcomes Study Short Form 12 (SF-12 ver-
sion 1) was used to measure HRQoL in the past 4 weeks.
The 12 questions in this instruments were combined and
transformed into two different 0–100 scales (higher values
indicating a better HRQoL), representing the physical and
mental components [20, 21]. This is a widely used instru-
ment with robust psychometric properties [8, 20].
Cardiometabolic conditions
Obesity was defined as a body mass index (BMI) ≥30 kg/
m2, based on self-reported weight and height, while the
other conditions were based on the medical diagnosis and/
or treatment for the specific disease. The CMRF for CVD
(obesity, diabetes, hypertension, and dyslipidaemia) were
then summed, and subsequently transformed into a binary
variable (positive when at least one of these conditions was
present). A similar procedure was used to combine the five
CVD (myocardial infarction, angina, heart failure, atrial
fibrillation/arrhythmia, and stroke). The resultant binary
variables were then used to create an ordinal one: (1) none
of these conditions; (2) just with CMRF, and; (3) with
CVD (with or without CMRF).
Covariates
Sociodemographic characteristics, lifestyle, and the use of
preventive medications were considered as covariates in
the analyses (4–6, 17, 20). Demographic variables included
gender (male or female), age (35–49; 50–64; 65–79;
≥80  years), marital status [married/living with a partner
or unmarried (never married, divorced, separated, or wid-
owed)], and residence area (urban or rural). Individual-
level SEP was operationalized by dwelling (owner; rented
privately; government/community housing) and attained
educational level (bachelor or higher; trade qualification;
certificate/diploma; secondary; less than secondary), while
macro-level SEP was investigated by the SEIFA-IRSAD
(in quartiles). Working status was investigated as employed
full-time, employed part-time, not working (unemployed,
home duties, and not working because work-related injury
or disability), or retired.
Four modifiable lifestyle risk factors were investigated:
(1) fruit/vegetable intake (consumption of 5+ portions of
fruit and/or vegetables/day); (2) physical activity level
(undertaking 30+ minutes of moderate/vigorous intensity
physical activity on 5+ days/week); (3) alcohol consump-
tion (0–2 standard drinks/day), and; (4) smoking status
(never, former, or current smoker). The use of preventive
medications for cardiometabolic conditions (antihyperten-
sive drugs, anti-diabetic medication, statins, and antiplate-
let agents) were investigated separately as binary (yes/no
questions) and combined into an ordinal variable (current
use of 0–1, 2, or 3+ of these medications).
Data analysis
All analyses considered the sampling design (clusters) and
were weighted considering the inverse of the individual´s
probability of selection within the household and re-
weighted to the estimated resident population in SA in
2014 [17, 18]. Absolute and relative frequencies (%) were
used to describe categorical variables, while mean and
standard deviation or median with interquartile range
(p25–p75) were used for numerical variables, depending
on the normality of the distribution. Confidence intervals of
95% (95%CI) were also estimated. Chi-square test for het-
erogeneity with Rao-Scott correction was used to verify the
association between sociodemographic variables and the
prevalence of cardiometabolic conditions.
Multiple linear regression models were used to evalu-
ate the association between cardiometabolic conditions and
HRQoL, as the physical and mental scores were normally
distributed. Three different analytic models were consid-
ered for adjustment in this study. In Model 1, the associa-
tion between cardiometabolic conditions and HRQoL were
adjusted for sociodemographic confounders. In Model 2,
lifestyle variables and the use of preventive medication
were additionally incorporated. Finally, for testing the mod-
erating role of sociodemographic characteristics, multipli-
cative interaction terms between each sociodemographic
13

variable and the presence of cardiometabolic conditions
were additionally incorporated in Model 3. In Models 1
and 2, variables were selected using a forward selection
process, considering a p value <0.10 in the bivariate asso-
ciation with HRQoL (Supplementary Tables S1 and S2).
A p value <0.10 for the interaction terms was also consid-
ered as indicative of effect modification in Model 3 [22].
Wald tests for heterogeneity or trend were used depending
on the nature of the independent variables. An alpha of 5%
was defined as indicative of statistical significance. Results
from crude and adjusted analyses were presented as regres-
sion coefficients (β) with their respective 95%CI. When
identified an effect modification, predicted adjusted means
of HRQoL in each category of the cardiometabolic variable
were estimated and presented graphically, stratified by the
respective sociodemographic moderator.
The residuals of final models were examined for normal-
ity and homoscedasticity to assess the adequacy of the lin-
ear regressions. Determination coefficients (r2) were used
to evaluate the overall model fit and the impact of the car-
diometabolic conditions on the HRQoL. Finally, the vari-
ance inflation factor (VIF) was investigated as an indicator
of possible collinearity between the explanatory variables.
All the analyses were performed in the version 14.0 of the
STATA statistical software (StataCorp, TX, USA).
This study was approved by the University of Adelaide
Human Research Ethics Committee (project H-097-2010).
Participants provided verbal rather than written informed
consent, due to the practicalities of carrying out a large
scale survey and the low risk nature of the survey content.
Results
The study included a sample of 2379 individuals aged 35+
years (mean age 57.1 ± 14.0 years; 51.7% females). Table 1
summarises the prevalence of cardiometabolic conditions
and distribution of HRQoL. Obesity and hypertension were
the most common CMRF, with 54.6% (95%CI 52.3–56.9)
having at least one of these conditions. Myocardial infarc-
tion was the most prevalent CVD and 13.0% (95%CI
11.7–14.4) of the sample were affected by at least one
CVD. The frequency of CVD increased with the number
of CMRFs: 3.9% for those without any CMRF, 11.0% for
those with 1 CMRF, 25.9% for those with 2 CMRF, and
37.3% for those with 3–4 CMRF (p value for trend <0.001;
data not shown in table). Except from a higher frequency
of dyslipidaemia in males, the prevalence and the number
of CMRF were not different according to gender. How-
ever, the prevalence of myocardial infarction, heart failure,
and the number of CVD were at least twice as higher in
males than in females. On the other hand, the mean of the
13
Qual Life Res
physical HRQoL was no different according to gender, but
the mental domain was slightly lower in females.
Table  2 shows most of the participants were married
(73.5%), living in urban areas (73.1%), household own-
ers (75.8%), had higher than secondary educational level
(65.4%), and employed full or part-time (53.1%). The prev-
alence of CMRF was slightly higher in females, but CVD
was twice as high in males (Table  2). Unmarried people,
those living in rural areas, and elderly people showed a
higher prevalence of both CMRF and CVD. Individuals
living in government/community housing had twice the
prevalence of CVD than those renting or household own-
ers. An educational level up to secondary school was asso-
ciated with a higher prevalence of CMRF, while CVD was
more frequent among those with less than secondary edu-
cational level or trade qualification. Retired individuals had
at least a three times higher prevalence of CVD than other
working status groups, but the differences in the prevalence
of CMRF factors were less marked. On the other hand,
CMRF and CVD showed an inverse trend association with
the SEIFA-IRSAD quartiles. Of the lifestyle variables, only
the practice of physical activity was associated with the
prevalence of cardiometabolic conditions (higher frequency
of CMRF and CVD those with lower physical activity
level). Only 70.7% of individuals with CVD used 3+ of
the preventive medications (antihypertensive, anti-diabetic,
statins, and/or antiplatelet agents).
Table  3 shows the crude and adjusted analyses for the
association between the cardiometabolic conditions and the
physical and mental components of HRQoL. In crude anal-
ysis, an inverse trend association was observed between the
occurrence of cardiometabolic conditions and the physi-
cal HRQoL, with a score 11.7 points lower among those
with CVD compared to healthy individuals. Adjustment for
sociodemographic confounders (Model 1) reduced these
differences by 33%. The adjusted r2 increased from 19.5%
(including just sociodemographic variables) to 24.0% when
the cardiometabolic conditions were included in Model 1.
However, the effect magnitude of the associations and the
r2 remained relatively stable after the inclusion of lifestyle
variables and use of preventive medication (Model 2). In
contrast, having a CMRF or a CVD was not associated with
the mental HRQoL in crude analysis or after adjustment
for sociodemographic variables (Model 1). In this case,
the r2 remained stable at 12.8%, even after the inclusion of
the cardiometabolic conditions in the model. When these
results were further adjusted for lifestyle variables and use
of preventive medication (Model 2), a slight inverse rela-
tionship became apparent (r2 = 14.7%).
When interaction terms between each sociodemo-
graphic variable and the cardiometabolic conditions were
incorporated in the analysis (Model 3), only the educa-
tional level and working status were effect modifiers of
Qual Life Res

Table 1  Descriptive Prevalence p value*

analysis for the prevalence of
cardiometabolic conditions and % 95%CI Gender
health-related quality of life
among individuals ≥35 years Males Females
in South Australia in 2015 % %
according to gender (N = 2379)
Cardiometabolic risk factors
 Obesitya 29.1 27.0;31.2 28.5 29.6 0.634
b
 Hypertension 30.4 28.4;32.4 31.6 29.3 0.269
 Dyslipidaemiab 22.9 21.1;24.7 25.6 20.1 0.004
 Diabetes ­mellitusb 11.8 10.6;13.3 12.6 11.2 0.313
 No. of cardiometabolic risk factors
  0 45.4 43.1;47.7 44.1 46.7 0.232
  1 27.7 25.7;29.8 26.9 28.4
  2 16.6 15.0;18.2 17.9 15.3
  3+ 10.3 9.1;11.7 11.2 9.6
Cardiovascular disease
 Myocardial infarction 7.0 6.0;8.1 10.8 3.4 <0.001
 Angina 3.1 2.4;3.8 3.7 2.5 0.081
 Atrial fibrillation/arrhythmia 4.1 3.4;5.0 4.5 3.8 0.844
 Stroke 2.3 1.8;3.0 2.7 2.0 0.217
 Heart failure 1.3 0.1;1.8 1.7 0.8 0.047
 No. of cardiovascular diseases
  0 87.0 85.6;88.3 82.9 90.8 <0.001
  1 9.6 8.5;10.9 12.6 6.8
  2 2.2 1.7;2.9 2.9 1.5
  3+ 1.2 0.8;1.7 1.6 0.8
Health-related quality of life
 Physical score (Mean ± SD) 46.8 ± 11.0 46.2;11.0 47.2 ± 9.9 46.5 ± 12.0 0.125**
 Mental score (Mean ± SD) 52.5 ± 9.1 52.0;53.0 53.0 ± 7.8 51.9 ± 10.2 0.003**

*Chi-square test for heterogeneity with Rao-Scott correction

**Wald test for heterogeneity
a
 Body mass index ≥30.0 kg/m², based on self-reported information on weight and height
b
 Medical diagnosis and/or using medication for that condition

the association with the physical component of HRQoL
(p value for heterogeneity <0.001 in both cases;
r2 = 28.4%). Figure  1a shows the predicted adjusted
means of the physical HRQoL according to the pres-
ence of cardiometabolic conditions, stratified by educa-
tional level. The physical HRQoL score among healthy
individuals was similar in all strata of educational level.
Nevertheless, when CVD were present, there was a lower
impact on the physical HRQoL among those with bach-
elor/higher educational level and trade qualification (2.6
and 4.1 lower HRQoL among those with CVD, respec-
tively). The impact of CVD were greater among those
with certificate/diploma, secondary, or less than second-
ary educational level: those affected by CVD had between
6.6 and 9.4 lower physical HRQoL compared with the
healthy ones. Figure  1b is consistent to show the mean
physical HRQoL score was similar in healthy individuals
in all strata of working status. Among employed (full or
part-time), having a CMRF decreased this score by two
points, doubling these differences when a CVD was pre-
sent. Among retired individuals, the presence of CMRF
was associated with a 3.1 lower HRQoL score, while
among those with CVD the difference was 7.5 points.
Nevertheless, among those classified as “not working”,
having a CMRF had the same impact on the HRQoL than
a CVD (9.7 point lower score when compared to those
free of these conditions). There was no evidence of het-
erogeneity in the associations between cardiometabolic
conditions and the mental HRQoL according to any of
the sociodemographic conditions (p value for interaction
>0.15 in all cases).
The mean VIF for both outcomes was 1.7, indicating no
collinearity between the explanatory variables in the final
models, and the residuals were normally distributed.

13
 Qual Life Res

Table 2  Bivariate association between sociodemographic, lifestyle, Table 2  (continued)

and use of preventive medications with the prevalence of cardiometa-
bolic conditions among individuals ≥35 years in South Australia, Variables % Cardiometabolic condition p value*
2015 (N = 2379) None Cardio- Car-
metabolic diovascular
Variables % Cardiometabolic condition p value*
­riska ­diseaseb
None Cardio- Car-
metabolic diovascular  0–4 days/week 65.3 40.9 45.1 14.0 0.004
­riska ­diseaseb  5+ days/week 34.7 48.7 40.2 11.1
Alcohol consumption
Gender
 >2 doses/day 31.8 45.9 43.0 11.1 0.170
 Male 48.3 42.6 40.2 17.1 <0.001
 0–2 doses/day 68.2 42.6 43.6 13.9
 Female 51.7 44.5 46.3 9.2
Current smoker
Age group
 Yes 14.7 44.7 43.1 12.3 0.889
 35–49 years 33.8 65.1 32.2 2.7 <0.001
 No 85.3 43.4 43.5 13.2
 50–64 years 35.6 42.2 48.5 9.3
Use of preventive ­medicationd
 65–79 years 23.3 24.4 52.2 23.4
 0–1 78.7 55.4 39.8 4.8 <0.001
 ≥80 years 7.3 11.9 42.3 45.8
 2 13.5 0.0 72.6 27.4
Marital status
 3+ 7.8 0.0 29.4 70.7
 Married 73.5 45.7 42.2 12.1 0.002
Total 100.0 43.6 43.4 13.0
 Unmarried 26.6 37.9 46.7 15.5
Residence area *Chi-square test for heterogeneity with Rao-Scott correction
 Urban 73.1 46.6 40.9 12.5 <0.001 SEIFA-IRSAD—Socio-Economic Indexes for Areas Index of Rela-
 Rural 26.9 35.5 50.1 14.5 tive Socio-economic Advantage and Disadvantaged
a
Dwelling  Including obesity (BMI ≥30  kg/m2), hypertension, diabetes and/or
 Owner 75.8 44.0 43.3 12.7 <0.001 dyslipidaemia, but not a diagnosis of any cardiovascular disease
b
 Rented privately 17.3 49.8 39.4 10.8  Including heart attack, angina, heart failure, stroke and/or atrial
fibrillation, with or without cardiometabolic risk factors
 Government/ 6.9 23.9 54.2 21.9 c
community  30+ minutes of moderate/vigorous intensity physical activity
d
housing  Number of medications used for the prevention of cardiovascular
Educational level diseases: antihypertensive drugs, anti-diabetic medication, statins,
and/or antiplatelet agents
 Bachelor or 23.3 60.0 32.8 7.3 <0.001
higher
 Trade qualifica- 13.6 39.1 42.0 19.0
tion Discussion
 Certificate/ 28.5 47.7 42.7 9.6
diploma This study investigated the extent to which sociodemo-
 Secondary 23.1 35.6 51.5 12.9 graphic variables increase the impact of cardiometabolic
 Less than sec- 11.5 22.0 51.8 26.2 conditions on HRQoL in a representative sample of adults
ondary living in a country with a very high human development
Working status and lower inequality indexes [18]. Four conclusions can
 Employed full 33.2 57.1 38.5 4.4 <0.001 be highlighted based on our results. First, cardiometabolic
time
conditions were associated with a lower physical HRQoL,
 Employed part 19.9 54.4 39.6 6.0
time
but the association with the mental component was subtle.
 Not working 15.9 44.1 46.0 9.9
Second, the impact of these conditions on the HRQoL is
 Retired 31.0 22.0 49.7 28.3
cumulative, with CVD doubling the effect magnitude of
SEIFA-IRSAD (quartiles)
CMRF on the physical HRQoL. Third, the influence of the
 High 25.1 52.5 37.7 9.9 <0.001
cardiometabolic conditions on the physical HRQoL was
 Middle-high 24.7 45.7 43.6 10.7
stronger among those with a lower educational level (cer-
 Middle-low 22.1 45.0 43.1 11.9
tificate/diploma, secondary, or lower), as well as those not
 Low 28.1 32.8 48.5 18.7
in the labour force (retired or not working). Finally, lifestyle
Fruit/vegetable intake
and the use of preventive medication had a very small effect
 0–4 portions/day 65.8 44.2 42.4 13.5 0.283
on the association between cardiometabolic conditions and
 5+ portions/day 34.2 42.6 45.4 12.1
the physical HRQoL.
Physical ­activityc
In the last 20 years, improvements in the management of
acute CVD reduced the age-adjusted mortality rates related

13
Qual Life Res

Table 3  Crude and adjusted analyses of the association between cardiometabolic conditions and health-related quality of life (physical and men-
tal components) among individuals ≥35 years in South Australia, 2015 (N = 2379)
Mean (95%CI) Crude Model 1 Model 2
β(95%CI) β(95%CI) β(95%CI)

Physical component
 Cardiometabolic condition p < 0.001* p < 0.001* p < 0.001*
  None 50.8 (50.1;51.5) Ref Ref Ref
  Cardiometabolic ­riska 45.1 (44.3;45.9) −5.6 (−6.6;−4.7) −3.7 (−4.8;−2.6) −3.6 (−4.8;−2.4)
  Cardiovascular ­diseaseb 39.1 (37.7;40.5) −11.7 (−13.2;−10.2) −7.8 (−9.5;−6.1) −7.2 (−9.1;−5.3)
Mental component
 Cardiometabolic condition p = 0.856* p = 0.085* p = 0.030*
  None 52.6 (52.0;53.2) Ref Ref Ref
  Cardiometabolic ­riska 52.2 (51.5;52.9) −0.4 (−1.3; 0.5) −0.6 (−1.4;0.2) −0.8 (−1.7;0.1)
  Cardiovascular ­diseaseb 52.8 (51.7;53.8) 0.2 (−1.0;1.3) −1.0 (−2.2;0.2) −1.6 (−3.0;−0.1)

*Wald-test for trend

Model 1—Adjusted for sociodemographic variables. For the physical component: age, marital status, dwelling, educational level, working sta-
tus, and SEIFA-IRSAD. For the mental component: gender, age, marital status, residence area, dwelling, educational level, working status, and
SEIFA-IRSAD
Model 2—Adjusted for variables in Model 1+ lifestyle variables (fruit/vegetable intake, physical activity level, alcohol consumption, and smok-
ing) and for the use of preventive medication for cardiovascular diseases. Alcohol consumption was not included in the analysis of the mental
component of HRQoL, as the p-value for that association was >0.10
a
 Including obesity (BMI ≥30 kg/m2), hypertension, diabetes and/or dyslipidaemia, but not a diagnosis of any cardiovascular disease
b
 Including heart attack, angina, heart failure, stroke and/or atrial fibrillation, with or without metabolic risk factors

to these conditions among Australians, increasing their
prevalence at the population level [3, 5]. Concomitantly,
the frequency of CMRF (obesity, hypertension, diabetes,
and dyslipidaemia) have also risen at a rate of 1.5–3.3%
per year [23]. The increment of CVD and CMRF are prob-
ably responsible for the progressive decrease in the physi-
cal score of HRQoL in the last two decades among adults
in SA [49.8 (95%CI 49.4–50.2) in 1997 to 48.6 (95%CI
48.2–49.0) in 2015]. However, the mental component of
HRQoL has remained stable [24]. This pattern is consist-
ent with our results of an inverse association between car-
diometabolic conditions and the physical but not with the
mental HRQoL. Accordingly, all of the 27 studies included
in a systematic review published in 2004 [10] found an
inverse relationship between the number of medical chronic
conditions and the physical domain of HRQoL, while the
association with the mental domain was unclear.
The cumulative effect of the cardiometabolic conditions
on the physical component of HRQoL has also been dem-
onstrated in another study conducted in 2008–2009 with a
sample of 587 patients with CVD recruited from 24 gen-
eral practices in two Australian states [12]. In that study,
the physical HRQoL score showed a mean reduction of 1.3
points for each additional CMRF, while the mean effect of
an additional cardiac conditions reduced in 3.0 points this
score. Again, neither the CMRF nor the CVD were associ-
ated with the mental HRQoL.
Different literature reviews have shown that cardio-
metabolic conditions (CVD, hypertension, dyslipidaemia,
diabetes) [25–28] and other chronic diseases [12, 14, 15]
predominantly affect individuals from lower SEP and/or
lower educational level. These inequalities are present even
in countries with universal health coverage, affecting the
whole treatment pathway, including access to health ser-
vices, diagnostic procedures, secondary prevention, and
management of acute conditions [27, 29].
Furthermore, according to our results, educational level
and working status are not only confounders in the relation-
ship between cardiometabolic conditions and the physical
HRQoL, but a source of heterogeneity in the observed asso-
ciations. Even after controlling for the other sociodemo-
graphic variables, lifestyle, and use of preventive medica-
tions, a lower educational level and not being in the labour
force (retired or not working) increased the effect magni-
tude of the association between cardiometabolic conditions
and the physical HRQoL. A population-based study con-
ducted in Germany (N = 11,177) [14] also found the effect
of diabetes mellitus, hypertension, and CVD on HRQoL
was higher among individuals with lower educational level.
A similar result was observed in a National Health Survey
conducted in England (N = 26,104) [15], where the HRQoL
impairment associated with CMRF (obesity, hypertension
or diabetes) was more evident among those with lower
(compared with high) occupational status. However, in that

13
 Qual Life Res

Fig. 1  Predicted mean for the a None Cardiometabolic risk Cardiovascular disease

physical component of health
60 β= -2.6 β= -4.1 β= -9.4 β= -6.6 β= -8.1
related quality of life (SF-12)
(95%CI -6.5;1.3) (95%CI -8.0;-0.2) (95%CI -13.2;-5.6) (95%CI -10.6;-2.6) (95%CI -12.3;-3.9)
according to the presence of

Quality of life - Physical component (score)

cardiometabolic conditions,
50
stratified by educational level
(Fig. 1a) and working status
(Fig. 1b). Results are adjusted 40
for variables with a p value
<0.10 in the association with
the physical component of 30
HRQoL (age, marital status,
dwelling, educational level,
20
working status, SEIFA-IRSAD,
fruit/vegetable intake, physical
activity level, alcohol con- 10
sumption, smoking, and use
of preventive medications for
cardiovascular diseases) and 0
interaction terms between Bachelor/higher Trade qualification Certificate/diploma Secondary Less than
cardiometabolic conditions and secondary
education and cardiometabolic b
conditions and working status 60 β= -5.4 β= -3.5 β= -9.7 β= -7.5
(Model 3). Vertical lines in the (95%CI -10.8;0.0) (95%CI -8.6;1.6) (95%CI -15.0;-4.4) (95%CI -10.0;-5.0)
Quality of life - Physical component (score)

columns represent the 95%CI. p

value for heterogeneity <0.001 50
in both cases
40

30

20

10

0
Employed full-time Employed part-time Not working Retired

study, there was no evidence of interaction between occu- not reduce the impact of cardiometabolic conditions on
pation and the diagnosis of ischaemic heart disease. This HRQoL. According to the literature, other patient-centered
discrepancy could be related to the lone score generated interventions may promote better HRQoL, such as a higher
by the EuroQol five-dimensions questionnaire (EQ-5D) frequency of contacts with the health service, better control
used in that study [15], which does not allow differentiating of biomedical risk factors, improving medication adher-
effects of cardiometabolic conditions on the different com- ence, encouraging regular physical activity, and enhancing
ponents of HRQoL. health literacy [12, 16, 25]. Notably, different studies have
Various mechanisms could explain the interaction shown health literacy is not only an indicator of the capac-
between a lower SEP and the diagnosis of cardiometa- ity of the individual to understand health information, but
bolic conditions on different health outcomes. They include also a protective factor for further complications and mor-
lower access to health care services, inadequate treatment tality among patients with CVD [25, 30]. Furthermore, the
plans, greater difficulties in adopting healthy habits, poorer impact of a lower health literacy on the HRQoL has been
mental health, and higher frequency of complications after demonstrated to be equivalent to the diagnosis of a new
diagnosis [25, 27–29]. From a clinical perspective, better cardiac condition among patients with an established CVD,
management of these chronic conditions among individu- even after controlling for age, gender, educational level,
als with lower educational level and/or unemployed seems and other socioeconomic indicators [12]. To elucidate the
like the most appropriate intervention to reduce inequalities relevance of this and other possible interventions to reduce
in HRQoL [16, 25–28]. However, according to our results, inequalities in HRQoL, future studies with cohorts of
lifestyle variables and the use of preventive medication did patients with these chronic conditions would be required.

13
Qual Life Res
Although this study has important strengths (representa-
tive sample, quality control of interviewed-based assess-
ments, robust instrument to evaluate the outcome), some
limitations must be recognised. First, the cross-sectional
design does not allow evaluation of either the temporality
of the associations between the cardiometabolic conditions
and the HRQoL. Nevertheless, some inherent temporality
was presumed in the investigated associations, considering
the diagnosis of CVD and CMRF (both chronic conditions)
precedes the recall length of the HRQoL questions (last
4 weeks). Second, the severity of CVD and CMRF was not
evaluated, and some of them were probably underreported
among individuals with a lower SEP [25, 27]. However,
these possible sources of information bias are less likely to
explain our results, as they would reduce the effect magni-
tude of the associations.
This study demonstrates that educational level and work-
ing status interact with the occurrence of CMRF and CVD,
increasing the adverse effects of these cardiometabolic
conditions on the physical component of HRQoL. There-
fore, individuals with a lower SEP are not only affected
by an increased frequency of these chronic conditions, but
also by lower levels of physical HRQoL once these condi-
tions are diagnosed [14, 15]. Our findings are even more
relevant considering the patterns in inequalities regarding
morbidity and mortality in Australia—a high-income coun-
try with one of the highest human development indexes
worldwide—which have remained relatively stable in the
last decade [31]. Therefore, interdisciplinary approaches
are necessary to reduce these health inequalities, as well
as new studies performed to identify the most appropriate
targeted interventions for public health practice. Given our
results, we suggest health practitioners should be aware not
only of the clinical parameters of patients with CMRF and/
or CVD, but also of their educational level and working
status. Besides a bigger effect on the physical HRQoL, a
lower SEP compromise the management of these chronic
diseases, increasing health disparities and the probability of
more severe complications.
Acknowledgements  The author acknowledges the participants of
the 2015 Spring Health Omnibus Survey for their participation in this
study.
Compliance with ethical standards  lowitz, D. J., & Lim, W. K. (2015). Relationship between health-
Funding  D.A. González-Chica received a part Fellowship from the
NHMRC Centre of Research Excellence to Reduce Inequality in Heart
Disease to conduct this study.
Conflict of interest  All the authors declare they have no conflict of
interest.
Ethical approval  This study was approved by the University of
Adelaide Human Research Ethics Committee (project H-097-2010).
All procedures performed in studies involving human participants
were in accordance with the ethical standards of the institutional and/
or national research committee and with the 1964 Helsinki Declaration
and its later amendments or comparable ethical standards.
Informed consent  Participants provided verbal rather than written
informed consent, due to the practicalities of carrying out a large-scale
survey and the low-risk nature of the survey content.
References
1. Hunter, D. J., & Reddy, K. S. (2013). Noncommunicable
diseases. The New England Journal of Medicine, 369(14),
1336–1343.
2. WHO. (2011). WHO maps noncommunicable disease trends
in all countries: Country profiles on noncommunicable disease
trends in 193 countries. Central European Journal of Public
Health, 19(3), 130–138.
3. AIHW. (2014). Australian Institute of Health and Welfare. Cardi-
ovascular disease, diabetes and chronic kidney disease—Austral-
ian facts: Prevalence and incidence. Cardiovascular, diabetes and
chronic kidney disease. Series No. 2. Cat. No. CDK 2. Canberra:
AIHW. (Available at http://www.aihw.gov.au/WorkArea/Down-
loadAsset.aspx?id=60129549614) Accessed on 15 Mar 2016.
4. Eckel, R. H., & Cornier, M. A. (2014). Update on the NCEP
ATP-III emerging cardiometabolic risk factors. BMC Medicine,
12, 115.
5. Begg, S. J. (2014). Health in a ‘post-transition’ Australia: Add-
ing years to life or life to years? Australian health review: a pub-
lication of the Australian Hospital Association, 38(1), 1–5.
6. Chowdhury, R., Khan, H., Heydon, E., Shroufi, A., Fahimi, S.,
Moore, C., Stricker, B., Mendis, S., Hofman, A., Mant, J., &
Franco, O. H. (2013). Adherence to cardiovascular therapy: A
meta-analysis of prevalence and clinical consequences. Euro-
pean Heart Journal, 34(38), 2940–2948.
7. Conn, V. S., Ruppar, T. M., Maithe Enriquez, R., & Cooper, P.
S. (2016). Patient-centered outcomes of medication adherence
interventions: Systematic review and meta-analysis. Value in
Health: The Journal of the International Society for Pharmaco-
economics and Outcomes Research, 19(2), 277–285.
8. Mark, D. B. (2016). Assessing quality-of-life outcomes in car-
diovascular clinical research. Nat Rev Cardiol.
9. Bakas, T., McLennon, S. M., Carpenter, J. S., Buelow, J. M.,
Otte, J. L., Hanna, K. M., Ellett, M. L., Hadler, K. A., & Welch,
J. L. (2012). Systematic review of health-related quality of life
models. Health and Quality of Life Outcomes, 10, 134.
10. Fortin, M., Lapointe, L., Hudon, C., Vanasse, A., Ntetu, A. L., &
Maltais, D. (2004). Multimorbidity and quality of life in primary
care: A systematic review. Health and Quality of Life Outcomes,
2, 51.
11. Hutchinson, A. F., Graco, M., Rasekaba, T. M., Parikh, S., Ber-
related quality of life, comorbidities and acute health care utili-
sation, in adults with chronic conditions. Health and Quality of
Life Outcomes, 13, 69.
12. Gonzalez-Chica, D. A., Mnisi, Z., Avery, J., Duszynski, K.,
Doust, J., Tideman, P., Murphy, A., Burgess, J., Beilby, J., &
Stocks, N. (2016). Effect of health literacy on quality of life
amongst patients with Ischaemic heart disease in Australian
General Practice. PLoS One, 11(3), e0151079.
13. Mello Ade, C., Engstrom, E. M., & Alves, L. C. (2014). Health-
related and socio-demographic factors associated with frailty in
the elderly: A systematic literature review. Cadernos de saude
13

publica/Ministerio da Saude, Fundacao Oswaldo Cruz, Escola
Nacional de Saude Publica, 30(6), 1143–1168.
14. Mielck, A., Vogelmann, M., & Leidl, R. (2014). Health-related
quality of life and socioeconomic status: Inequalities among
adults with a chronic disease. Health and Quality of Life Out-
comes, 12, 58.
15. Stafford, M., Soljak, M., Pledge, V., & Mindell, J. (2012).
Socio-economic differences in the health-related quality of life
impact of cardiovascular conditions. European Journal of Public
Health, 22(3), 301–305.
16. Ludt, S., Wensing, M., Szecsenyi, J., van Lieshout, J., Rochon,
J., Freund, T., Campbell, S. M., & Ose, D. (2011). Predictors of
health-related quality of life in patients at risk for cardiovascular
disease in European primary care. PLoS One, 6(12), e29334.
17. Taylor, A., Dal Grande, E., & Wilson, D. (2006). The South Aus-
tralian Health Omnibus Survey 15 years on: has public health
benefited? Public Health Bull (S Aust), 3(1), 30–32. Available
at http://pandora.nla.gov.au/pan/133553/20120522-0000/www.
sahealth.sa.gov.au/wps/wcm/connect/9d76de80440e1c688bc8af
63794072bf/phb-chronicdisease065ef3.pdf. Accessed on 16 Jun
2016.
18. ABS. (2016). Australian Bureau of Statistics. Table Builder.
Available at http://www.abs.gov.au/websitedbs/censushome.nsf/
home/tablebuilder. Accessed on 10 May 2016.
19. ABS. (2011). Australian Bureau of Statistics. Census of Popu-
lation and Housing: Socio-economic indexes for areas (SEIFA),
Australia. Available at http://www.abs.gov.au/ausstats/abs@.
nsf/mf/2033.0.55.001. Accessed on 01 Mar 2014 (Vol. cat. no.
2033.0.55.001).
20. Gandek, B., Ware, J. E., Aaronson, N. K., Apolone, G., Bjorner,
J. B., Brazier, J. E., Bullinger, M., Kaasa, S., Leplege, A., Prieto,
L., & Sullivan, M. (1998). Cross-validation of item selection and
scoring for the SF-12 Health Survey in nine countries: Results
from the IQOLA Project. International Quality of Life Assess-
ment. Journal of Clinical Epidemiology, 51(11), 1171–1178
21. Wilson, D., Tucker, G., & Chittleborough, C. (2002). Rethinking
and rescoring the SF-12. Sozial- und Präventivmedizin, 47(3),
172–177.
22. Mitchel, M. (2012). Interpreting and visualizing regression mod-
els using STATA, first ed (First ed.). Texas:Stata Press.
13
Qual Life Res
23. Taylor, A. W., Dal Grande, E., Wu, J., Shi, Z., & Campostrini,
S. (2014). Ten-year trends in major lifestyle risk factors using an
ongoing population surveillance system in Australia. Population
Health Metrics, 12(1), 31
24. Avery, J., Dal Grande, E., & Taylor, A. (2004). Quality of life in
South Australia as measured by the SF-12 Health Status Ques-
tionnaire: Population norms for 2003: Trends from 1997 to 2003.
ISBN 0730893294. Available at http://www.health.adelaide.
edu.au/pros/docs/reports/general/qol_quality_of_life_sf_12.pdf.
Accessed on 05 Jun 2015 (No. .). South Australia: South Aus-
tralia. Dept. of Human Services. Population Research and Out-
come Studies Unit.
25. Diaz-Toro, F., Verdejo, H. E., & Castro, P. F. (2015). Socioeco-
nomic inequalities in heart failure. Heart Failure Clinics, 11(4),
507–513.
26. Hawkins, N. M., Jhund, P. S., McMurray, J. J., & Capewell, S.
(2012). Heart failure and socioeconomic status: Accumulat-
ing evidence of inequality. European Journal of Heart Failure:
Journal of the Working Group on Heart Failure of the European
Society of Cardiology, 14(2), 138–146.
27. Schroder, S. L., Richter, M., Schroder, J., Frantz, S., & Fink, A.
(2016). Socioeconomic inequalities in access to treatment for
coronary heart disease: A systematic review. International Jour-
nal of Cardiology, 219, 70–78.
28. Sommer, I., Griebler, U., Mahlknecht, P., Thaler, K., Bouskill,
K., Gartlehner, G., & Mendis, S. (2015). Socioeconomic ine-
qualities in non-communicable diseases and their risk factors:
An overview of systematic reviews. BMC Public Health, 15,
914.
29. Terraneo, M. (2015). Inequities in health care utilization by peo-
ple aged 50+: Evidence from 12 European countries. Social Sci-
ence & Medicine (1982), 126, 154–163.
30. Cajita, M. I., Cajita, T. R., & Han, H. R. (2016). Health literacy
and heart failure: A systematic review. The Journal of Cardio-
vascular Nursing, 31(2), 121–130.
31. AIHW. (2014). Australian Institute of Health and Welfare. Mor-
tality inequalities in Australia 2009–2011. Bulletin no. 124. Cat.
no. AUS 184. Canberra: AIHW. Available at http://www.aihw.
gov.au/publication-detail/?id=60129548021. Accessed on 03
Mar 2016.