Archives of Gerontology and Geriatrics 70 (2017) 123–129

Contents lists available at ScienceDirect

Archives of Gerontology and Geriatrics

journal homepage: www.elsevier.com/locate/archger

A structural equation model of the relation between socioeconomic

status, physical activity level, independence and health status in older
Iranian people
Zahra Mosallanezhada,b,c,* , Gholam Reza Sotoudehd,e , Göran Jutengrenf ,
Mahyar Salavatia , Karin Harms-Ringdahlb , Lena Nilsson Wikmarb,h , Kerstin Frändinb,g
a
Department of Physiotherapy, University of Social Welfare and Rehabilitation Sciences, Tehran, Iran
b
Division of Physiotherapy, Department of Neurobiology, Care Sciences and Society, Karolinska Institutet, Stockholm, Sweden
c
Iranian Research Centre on Aging, University of Social Welfare and Rehabilitation Sciences, Tehran, Iran
d
Department of Health Sciences, Mid Sweden University, Sundsvall, Sweden
e
Sina Trauma and Surgery Research Center (STSRC), Sina General Hospital, Tehran University of Medical Sciences, Tehran, Iran
f
Department of Work Life and Social Welfare, University of Borås, Sweden
g
Department of Neuropsychiatric Epidemiology, Institute of Neuroscience and Physiology, Sahlgrenska Academy, University of Gothenburg, Gothenburg,
Sweden
h
Academic Primary Healthcare Centre, Stockholm County Council, Stockholm, Sweden

A R T I C L E I N F O A B S T R A C T

Article history: Background and aim: Health status is an independent predictor of mortality, morbidity and functioning in
Received 8 May 2016 older people. The present study was designed to evaluate the link between socioeconomic status (SES),
Received in revised form 29 July 2016 physical activity (PA), independence (I) and the health status (HS) of older people in Iran, using structural
Accepted 9 January 2017
equation modelling.
Available online 19 January 2017
Methods: Using computerized randomly selection, a representative sample of 851 75-year-olds living in
Tehran (2007–2008), Iran, was included. Participants answered questions regarding indicators of HS, SES
Keywords:
and also PA and I through interviews. Both measurement and conceptual models of our hypotheses were
Old people
Structural equation model
tested using Mplus 5. Maximum-likelihood estimation with robust standard errors (MLR estimator), chi-
Health square tests, the goodness of fit index (and degrees of freedom), as well as the Comparative Fit Index (CFI),
Socio-economic status and the Root Mean Square Error of Approximation (RSMEA) were used to evaluate the model fit.
Physical activity Results: The measurement model yielded a reasonable fit to the data, x2 = 110.93, df = 38; CFI = 0.97;
Independence RMSEA = 0.047, with 90% C.I. = 0.037–0.058. The model fit for the conceptual model was acceptable;
ADL x2 = 271.64, df = 39; CFI = 0.91; RMSEA = 0.084, with 90% C.I. = 0.074–0.093. SES itself was not a direct
predictor of HS (b = 0.13, p = 0.059) but it was a predictor of HS either through affecting PA (b = 0.31,
p < 0.001) or I (b = 0.57, p < 0.001).
Conclusion: Socioeconomic status appeared to influence health status, not directly but through mediating
some behavioral and self-confidence aspects including physical activity and independence in ADL.
© 2017 Elsevier B.V. All rights reserved.

1. Introduction such as physical activity habits, disability, morbidity and mortality

The importance of self-assessed global health as an indepen-
dent predictor of mortality, morbidity and functioning in
community-living older people has been well documented
(Yunhwan, 2000). The fact that social and economic discrepancies
contribute to the unequal distribution of health-related behaviors
* Corresponding author at: Department of Physical Therapy, University of Social
Welfare and Rehabilitation Sciences, Tehran 1985713831, Iran.
E-mail address: zmosallanezhad@yahoo.com (Z. Mosallanezhad).
has also been demonstrated (Dragano et al., 2007; Stewart, 2005;
Umberson & Montez, 2010). The interaction of different dimen-
sions of socio-economic status (SES) (individual characteristics,
urban neighborhood, cities and countries) (Dragano et al., 2007;
Elovainio, Kivimäki, Kortteinen, & Tuomikoski, 2001) and cultural
resources (comprehension, attitudes and behavior) influence
health status (HS) (Umberson & Montez, 2010). It has been
pointed out that numerous intermediating factors are involved in
this link (Dragano et al., 2007).
Physical inactivity is among the important risk factors for
morbidity and mortality (Dragano et al., 2007; Yunhwan, 2000).

http://dx.doi.org/10.1016/j.archger.2017.01.004
0167-4943/© 2017 Elsevier B.V. All rights reserved.
124 Z. Mosallanezhad et al. / Archives of Gerontology and Geriatrics 70 (2017) 123–129

Sedentary behavior has been shown to be more common at old age to communicate) were excluded. Unintentionally, the sample
than earlier in life (Stewart, 2005; Umberson & Montez, 2010). The obtained for this study only included people living in private
encouragement of a physically active life style has therefore been households. The final study sample numbered 851 subjects, 395
introduced as a feasible way to promote the psychological and women and 456 men (Fig. 1).
physical health of the general public (Brown et al., 2003). An
individual’s physical activity habits are shaped throughout life, and 2.2. Procedure
various factors, including socioeconomic and cultural determi-
nants as well as environmental and social burdens, influence his/ The participants were invited to take part in interviews and test
her habits in this respect (Watt, Carson, Lawlor, Patel, & Ebrahim, sessions by telephone. Certain questions from the interviews were
2009). Independence in activities of daily living (ADL) has been used in the present study (Mosallanezhad, Hörder, Salavati, Nilsson
also shown to affect health (Lee & Jeon, 2005). It may vary Wikmar, & Frändin, 2012). Instruments and questionnaires that
according to a person’s socioeconomic status; an old person who is had been used in the Swedish study, one of the longitudinal
independent in ADL may value her/his health status highly and try Gerontological and Geriatric Population Studies in Gothenburg,
to improve it (Lee & Jeon, 2005). H70 (Eriksson, Mellström, & Svanborg, 1987), were selected. A
Building and testing various models to show the intermediating standard method of forward/backward translation was used to
links to HS can help to discover the influencing factors (Dowd & provide a Persian version (Mosallanezhad et al., 2011). In addition,
Goldman, 2006; Kıvanç, Akova-Budak, Olcaysü, & Çevik, 2016). some questions from a Persian study addressing aspects of physical
Such models can illuminate the direction and degree of direct and and mental health as well as socio-economic status based on
indirect effects of health-related determinants (Dowd & Goldman, material from the municipality database were used. All interviews
2006; Elovainio et al., 2001; Yunhwan, 2000). During the last thirty were carried out by the author. Raw data were considered to be
years, methods and software have been used for structural observable variables and a set of observed variables was classified
equation modeling (SEM). SEM allows researchers in different for each latent variable reflecting the concepts of HS, SES, PAL and I.
fields to assess their theories empirically. These theories are
usually formulated as theoretical models for observed and latent 2.3. Data collection
(unobservable) variables. If data are collected for the observed
variables of the theoretical model, proper programs such as Mplus 2.3.1. Indicators of health status (HS)
5 can be used to test whether the model fits the data. Mplus can Participants indicated if they felt healthy (yes or no), generally
estimate both structural equation models and path models for a tired (yes or no) and rated their physical fitness as being either very
single or multiple groups. In addition, it can estimate models with poor, poor, quite good, good or very good (Mosallanezhad et al.,
regressions among combinations of continuous latent variables 2012).
and observed variables. In Mplus, both factor indicators and other
observed dependent variables for these models can be continuous,
censored, binary, ordered categorical (ordinal), counts, or combi-
nations of these variable types (Muthén & Muthén, 2007).
Considering the results of previous studies (Lee & Jeon, 2005;
Pruchno, Wilson-Genderson, & Cartwright, 2012; Schöllgen,
Huxhold, Schüz, & Tesch-Römer, 2011; Tse, Rochelle, & Cheung,
2011), we hypothesized that SES can affect HS in different ways;
either directly or indirectly by mediating factors such as physical
activity level (PAL) and independence (I). Mediating factors can
explain or modulate the direct pathway. Many studies in this field
have failed to include a representative sample of community-
dwelling people, and most of them describe Western countries
(Braveman, 2011; Cockerham, 2005; Tse et al., 2011), where life
style, cultural norms concerning social relationships and SES differ
from this in a country like Iran. Although it is possible that the SES
gradients in health observed in Iran are similar to those in other
countries, the pathways between SES and health may be different.
The aim of this study was to develop a confirmatory structural
equation model to test the links between the concepts socioeco-
nomic status, physical activity, independence and health status in
75-year-olds in Tehran.

2. Methods

2.1. Participants

In a cross-sectional study, a representative sample of 75-year-

olds born 1932–1933 and living in Tehran, Iran, 2007–2008 was
included by randomly selecting 1100 subjects from the latest
Iranian census records (1996) by the Center of Statistics in Iran
using computerized methods. The sample size was based on an
expected drop-out rate of about 20%. All elderly people of the
randomly selected sample could be included in the study, but
subjects with severe functional disability (not able to fulfill the Fig. 1. Flow chart, indicating the recruitment procedure of the 75-year-old
most of functional tests) and/or communication deficits (not able participants in Tehran, Iran.
 Z. Mosallanezhad et al. / Archives of Gerontology and Geriatrics 70 (2017) 123–129 125

2.3.2. Indicators of socioeconomic status (SES)
Socioeconomic indicators included overall years of education,
job status: no formal job, being a worker (factory worker or
labourer), voluntary work, official jobs (municipal or state
employees), high ranked jobs (such as university professor or
physician) and monthly family income, where the last two
variables were from the Persian database (Mosallanezhad et al.,
2011; Teymoori & Foroozan, 2005).
2.3.3. Indicators of physical activity level (PAL)
It includes physical activity level in the summer and winter
seasons according to a six-grade scale, including household
activities, ranging from hardly any physical activity to hard
exercise several times/week. A high grade means being more
physically active, and the scale has been shown to be valid (Frändin
& Grimby, 1994).
2.3.4. Indicators of independence (I)
Using the ADL staircase (Mosallanezhad et al., 2012; Sonn,
Törnquist, & Svensson, 1999), participants rated their level of
dependence on another person and their feeling of un-safety when
doing five defined personal activities of daily living (P-ADL) and four
defined instrumental activities of daily living (I-ADL). Cumulative
scales, as number of dependent/un-safe activities, were used in the
data analysis, where high scores mean dependent/un-safe in more
activities. Answers to questions regarding independence in usual
tasks indoors obtained from the Persian database (Teymoori &
Foroozan, 2005) were also included.
errors (MLR estimator), which can effectively deal with the data
that are not normally distributed.
In principle, a non-significant chi-square test would signify that
the data provided a good fit to the model. This test can elucidate the
sum of differences between observed and expected outcome
frequencies. However, because the goodness of fit test is
problematic with large samples (Hayduk, 1996), the adequacy of
the models was described with some additional statistics, e.g. the
root mean square residuals and the adjusted goodness of fit.
Therefore, besides the goodness of fit index (and degrees of
freedom), we used the (Comparative Fit Index; CFI), and the Root
Mean Square Error of Approximation (RSMEA) to evaluate model
fit. The CFI (Bentler, 1990) can vary between 0 and 1 (higher values
indicate better model fit) and measures how well the model fits
relative to a baseline model (Bentler, 1990). As a common rule of
thumb, values greater than 0.95 indicate a good fit to the data, and
values greater than 0.90 indicate an acceptable fit (Hu & Bentler,
1999). The RMSEA (Browne & Cudeck, 1993) measures the amount
of discrepancy between a specified model and the collected data
(lower values indicate better model fit) (Browne & Cudeck, 1993). It
has been suggested that a value of 0.05 and below indicates a good
fit to the data, a value of about 0.08 and less indicates a sensible fit,
and that values greater than 0.10 should not be accepted (Browne &
Cudeck, 1993).
3. Results
3.1. Participants

2.4. Ethics
The study was approved by the Ethics Committee of the
Ministry of Health, Treatment and Medical Education, Tehran, Iran.
Informed consent was obtained from each participant before the
interviews and tests were held.
2.5. Data analysis
A hypothesized model including the concepts (as latent
variables) and indicators (as the observable variables) was
designed. Mplus 5 was used to test the structural equation model
of our hypotheses (Muthén & Muthén, 2007). The analysis was
performed in two steps. In the first step, we tested a measurement
model to establish if we had chosen relevant measures to indicate
each of the latent variables. Testing the measurement model
involved relating the observed variables to the underlying concepts
by means of confirmatory factor analysis. In the second step, our
conceptual model was tested to evaluate the hypothesized links
between the latent variables (HS, SES, PAL and I). In both steps, we
used the maximum-likelihood estimation with robust standard
About half of the participants were women, and 51.6% of the
total population reported feeling healthy. Overall years of
education and monthly family income were significantly higher
for men than for women (p < 0.05). Of the women, 86.1% were
housewives. Of the men, 46.5% had voluntary work, 36% official
jobs, 10.7% were workers and 5.3% had a high ranked job. The most
frequent physical activity level was 3 for both summer and winter
seasons (more than 40%), and half of the participants rated their
physical fitness as quite good. Results of the cumulative scales
regarding the ADL staircase (Sonn et al., 1999) showed that 45% of
the participants were independent and 48.8% felt safe when doing
ADL. In response to the question regarding independence in usual
tasks indoors, 56.6% answered yes. Bivariate correlations and mean
values of the study variables are shown in Table 1. Men showed
better results than women regarding the indicators of PAL and I
(p < 0.05).
3.2. Measurement model
To verify if HS, SES, PAL and I can be preserved as separate latent
constructs, a measurement model was tested (Fig. 2). The model

Table 1
Bivariate correlations, mean values (M) and standard deviations (SD) of study variables (N = 851).

Variable M SD 1 2 3 4 5 6 7 8 9 10
1. Summer Physical Activity level 2.22 0.68
2. Winter Physical Activity Level 3.11 0.79 0.45
3. Feeling healthy 2.90 0.73 0.28 0.29
4. Feeling generally tired 1.81 0.77 0.17 0.34 0.39
5. Physical fitness 4.54 0.78 0.29 0.26 0.60 0.29
6. ADL Staircase, (Dependent in  1 activity) 3.55 0.78 0.18 0.30 0.29 0.53 0.33
7. ADL Staircase, (Unsecure in  1 activity) 2.91 0.79 0.21 0.14 0.16 0.14 0.17 0.12
8. Independence in usual tasks indoor 3.55 0.82 0.20 0.30 0.24 0.28 0.21 0.20 0.56
9. Overall years of education 2.92 0.68 0.20 0.17 0.12 0.15 0.15 0.20 0.58 0.40
10. Job status 3.17 0.50 0.56 0.42 0.31 0.22 0.45 0.11 0.18 0.32 0.18
11. Mean of the monthly family income 2.64 0.68 0.20 0.39 0.20 0.25 0.20 0.26 0.37 0.64 0.43 0.37

Note: All coefficients are significant at p < 0.05.

126 Z. Mosallanezhad et al. / Archives of Gerontology and Geriatrics 70 (2017) 123–129

Fig. 2. The measurement model confirms relations between observable variables with latent variables. Factor loadings for HS were 0.44 for SES, 0.51 for PAL and 0.77 for I. P-
values for estimates of relations between each latent variable with its related observable variables were equal to zero.
HS: Health status. FGT: Feeling of general tiredness. FH: Feeling healthy. PF: Physical fitness.
SES: Socio-economic status. MFI: Monthly family income. OYE: Overall years of education.
JS: Job status.
PAL: Physical Activity Level. WPA: Winter physical activity level. SPA: Summer physical activity level.
I: Independence. TA: Total A: ADL staircase, dependence (total score of 0: completely independent). TB: Total B: ADL staircase, safety (total score of 0: completely safe).
IIA: Independence in indoor activities.

also included two-way paths between the concurrent latent 4. Discussion

constructs. This model yielded a reasonable fit to the data,
x2 = 110.93, df = 38; CFI = 0.97; RMSEA = 0.047, with 90% C.
I. = 0.037–0.058.
3.3. Conceptual model
A proposed structural equation model was designed. It reflected
the impact of SES diversity on the health status of older persons,
relations between the determinants of each variable, as well as the
probable links between independent and dependent variables
(Figs. 2 and 3).
The cross-sectional inter-correlations between SES, PAL, I and
HS were tested. The estimation of this hypothesized structural
model yielded an acceptable fit to the data, x2 = 271.64, df = 39;
CFI = 0.91; RMSEA = 0.084, with 90% C.I. = 0.074–0.093. The con-
ceptual links are displayed in Fig. 3. As the figure shows, SES itself
was not a direct predictor of HS (b = 0.13, p = 0.059), but it was
predictor of HS either through affecting PAL (b = 0.31, p = 0.000) or I
(b = 0.57, p = 0.000). And both these factors could influence HS
positively.
The model fit for the conceptual model was accepTable SES was
not a direct predictor of HS but it could influence HS through
intermediating between PAL and I. SES itself could affect PAL and I
positively, meaning that participants with a high SES had a higher
physical activity level and were more independent in ADL than
others. A high SES can support the possibility of a high physical
activity level and independence and, as a consequence, result in
better health.
The present model included some socioeconomic, behavioral
and self-confidence aspects. In addition to the factors we included,
there are many other intrinsic and extrinsic factors such as
genetics, nutrition and sleeping habits, emotional and spiritual
aspects, culture, environmental factors and access to health
services that can affect the links to HS (Brown et al., 2003;
Dragano et al., 2007; Umberson & Montez, 2010; Watt et al., 2009;
Dowd & Goldman, 2006). More research is needed to clarify the
mechanisms behind the link between SES and HS as well as the
possible intermediating pathways. Differences in access to medical
care, health-related behavior, psychosocial stress and more
 Z. Mosallanezhad et al. / Archives of Gerontology and Geriatrics 70 (2017) 123–129
Fig. 3. Conceptual model for the whole population (a), shows relations between socio-economic status (SES) and health status (HS) and the intermediating role of physical
activity level (PAL) and independence (I).
exposure of persons with a low SES to life stressors (Dowd &
Goldman, 2006; Akova-Budak, Olcaysü & Çevik, 2016; Iversen &
Kraft, 2006), their weaker psychological coping strategies, more
hostility and less perceived control (Elovainio et al., 2001; Dowd &
Goldman, 2006; Pruchno et al., 2012) have been proposed as
involved trails. A new concept of the cultural capital offered by
recent studies explains not only how cultural, social and economic
reserves organize people’s behavioral options and preferences,
health-related values and norms but also their perceptions and
knowledge about a healthy lifestyle (Umberson & Montez, 2010). It
can therefore illuminate the related process leading to inequalities
in health.
Previous studies have also shown the link between better
general health and higher income, more education, and current
employment (Elovainio et al., 2001; Rutt & Karen, 2005; Sun et al.,
2007). Other findings identified that people in the lower
socioeconomic positions worried more about their health, had
higher treatment expenditure and evaluated their health as poorer
in comparison to the higher SES group (Abbema, Van Assema, Kok,
De Leeuw, & De Vries, 2004; Park, Sohn, Lee, & Kwon, 2014). But
conflicting results have been obtained from studies focused on
health determinants and health-related risk factors among people
in different socioeconomic positions. These results ranged from a
strong negative link between SES and health-related risk factors
(Iversen & Kraft, 2006; Lantz et al., 2001; Rodriguez et al., 2004), no
difference (Lee & Jeon, 2005; Pruchno et al., 2012), and a positive
link (Rutt & Karen, 2005). This discrepancy may be related to the
different approaches of the studies, including design, methodology
and sample size, or intraindividual, cultural and environmental
differences (Lee & Jeon, 2005; Rodriguez et al., 2004).
The present study emphasized the role of physical activity and
independence as the predictors of health. According to both
longitudinal and cross-sectional studies, the absence of health-risk
behaviors such as physical inactivity can postpone and reduce the
risk of disability, morbidity and mortality (Chakravarty et al., 2012;
Hillsdon, Lawlor, Ebrahim, & Morris, 2008). An effective health
intervention strategy to increase or maintain functional perfor-
mance and quality of life among older adults can involve
supporting physical activity habits (Chakravarty et al., 2012;
Fiatarone Singh, 2002; Schulz et al., 2008). A sedentary life style
adversely affects people's health status (Brodersen, Steptoe,
127
Boniface, & Wardle, 2007; Lantz et al., 2001; Pruchno et al.,
2012). A limited access to facilities for different types of exercise is
common in neighborhoods inhabited by people with a poor
education and low income. These limitations can lead to stressful
conditions and little time for exercise, resulting in a tendency
toward a low level of physical activity (Dragano et al., 2007;
Hillsdon et al., 2008; Lantz et al., 2001; Schulz et al., 2008). In
contrast, neighborhoods with a nicely planned environment and
accessibility to parks, gyms and other facilities for exercise have a
positive influence on health (Hillsdon et al., 2008; Maas, Verheij,
Spreeuwenberg, & Groenewegen, 2008).
Both material and non-material resources encourage healthy
lifestyle patterns (Hillsdon et al., 2008). Investigations have even
shown a tendency among physically active people to migrate into
supportive environments (Maas et al., 2008). Some studies have
also shown that the perception and knowledge about health-
related factors and social norms supporting physical activity and
access to physical activity resources vary in different SES
neighborhoods (Hanson & Chen, 2007; Maas et al., 2008;
Mackenbach & Bakker, 2003; Mulder, de Bruin, Schreurs, van
Ameijden, & van Woerkum, 2011; Wormald, Waters, Sleap, & Ingle,
2006;).
The association between decline in global ADL and indepen-
dence of elderly people has been documented (Lee & Jeon, 2005).
Studies confirm the relation between confidence in doing ADL and
functioning with better health in old adults (Lee & Jeon, 2005). In a
study on 796 elderly people, it is revealed that participants with
less physical activity, more diseases, and not participate in social
activities showed higher probability of being dependent in both
ADL and IADL (Kim et al., 2012). The association between need for
assistance for at least one ADL with age, low education, low
income, and impairments (motor, language, visual, mental and
other), has been shown, the researchers concluded that the
declared needs for human assistance are based not exclusively on
functional limitations but are also related to the social and
environmental setting (Bostan, Oberhauser, Stucki, Bickenbach, &
Cieza, 2015; Davin, Paraponaris, & Verger, 2005). These results
emphasis that functional independence and promoting healthy
aging are achievable through intervention programs promoting
healthy life-style and managing chronic diseases earlier in life and
facilitating social participation in later life.
128 Z. Mosallanezhad et al. / Archives of Gerontology and Geriatrics 70 (2017) 123–129
5. Limitations and suggestions
Various studies have emphasized the use of structural equation
modeling to determine the direction and significance of the links
between different constructs including socioeconomic and health
discrepancies (Stewart, 2005; Dowd & Goldman, 2006; Sun et al.,
2007). However, the models that were designed and tested in the
present study indicate a linear equation relationship between the
variables, which is a limitation. A more extensive linear or a non-
linear model including most influencing factors could result in a fit
model with more acceptable predictability. Since the present
model does not address the possibility of non-linear relationships,
further research is needed to establish the existence of such
relationships, including age (Hillsdon et al., 2008), psychological
and cultural (Elovainio et al., 2001; Pruchno et al., 2012; Iversen &
Kraft, 2006; Sun et al., 2007), as well as environmental, lifestyle
and other health-influencing determinants (Hillsdon et al., 2008;
Lantz et al., 2001). A multivariate model that includes all these
aspects would be a helpful tool for gaining a better understanding
of the role of each factor in the multifaceted pathways linking
unequal socioeconomic positions to inequalities in health (Maas
et al., 2008).
One of the strengths of the present study was the large,
representative sample of community- dwelling old people. There
are, however, some limitations regarding the methodology. In
general, path diagrams including Mplus play a fundamental role in
structural modeling, and they indicate the causal relationship
between independent and dependent variables that they are
interconnected. Longitudinal studies would be more appropriate
for establishing casual relationships (Elovainio et al., 2001;
Pruchno et al., 2012).
The present study focused on self-rated health which is
assumed to be a reliable tool for assessing health status. It can
also be a good predictor of functional ability in old people (Sun
et al., 2007; Yunhwan, 2000). But the use of self-reported data for
behavioral indicators could be considered a limitation. Some
studies indicated that individuals might over-report their physical
activity level (Hillsdon et al., 2008; Rutt & Karen, 2005). However,
previous studies have confirmed the homogeneity of the results
related to both subjective and objective variables (Mosallanezhad
et al., 2012). Monthly family income was also self- reported. Many
of the participants tended to report its approximate value. In
addition, as a historical tradition in the world and a common
cultural trend especially in the old population in Iran, the husband
is responsible for providing financial support to his wife and
children and takes care of most outdoor affairs including shopping.
The woman is mostly responsible for household activities and for
taking care of the children. Usually the man allocates money and
necessities to his wife and children based on the amount of his
earnings and their needs, something that may have affected the
participants’ description of monthly income. The main aim of the
present study was to describe the relations between the
determinants included. At this phase, we did not plan to evaluate
gender differences. In addition, structural models would run more
logical and consistent when a large sample is included (Boomsma,
1982; Gagné & Hancock, 2006; Velicer & Fava, 1998). Thus, a future
study is needed to run and test separate structural models for
women and men. It would help to better understand related
pathways.
6. Conclusion
Results of this study using structural equation modeling was
able to illuminate some intermediation paths between socioeco-
nomic and health status. The model fit for the conceptual model
was acceptable and confirmed that SES was not a direct predictor
of HS but it could influence HS through intermediating some
behavioral and self- confidence aspects including PAL and I. This
means that SES does not determine health, but that some factors
that are more prominent in people with a high SES, such as good
physical activity habits and independence in ADL, influence health
considerably. The participants with a high SES had a higher
physical activity level and were more independent in ADL.
Providing opportunities for a high physical activity level and
independence, a good SES could improve health.
7. Implications
 These findings can influence people’s belief and knowledge
about the factors influencing health. It can encourage them to
eliminate risky behaviors and improve their healthy and active
life style and the circumstances which can influence their health.
 The links elucidated by the present study and the models can
influence the government’s decision to address predisposing
factors in inequality with regard to health. Public policy and
health policy need to work together to inform one another, and
be directed toward countering the life circumstances and
behaviors that generate poor health and promoting those that
give rise to good health. They should also remove barriers to
increasing healthy, social and physical behaviors.
Declaration of interest
Authors have no conflict of interest.
Ethics
The study was approved by the Ethics Committee of the
Ministry of Health, Treatment and Medical Education, Tehran, Iran.
Acknowledgements
This study was part of an international study to evaluate the
physical activity patterns and functional performance in a group of
75-year-olds living in Iran and compare them to those of peers in
Sweden. The authors wish to thank both Gothenburg University
and Karolinska Institutet, Sweden, the Iran Ministry of Health,
Treatment and Medical Education, the Country Welfare Organiza-
tion, the Municipal Cultural Organization for Old People and the
Kahrizak Institute for the Elderly, Tehran, Iran for providing
scientific and financial support. We thank Göran Jutengren for
invaluable statistical support. We also thank all participants, Dr
Mohammad Taghi Joghatai, Dr. Sadat Seyed Bagher Maddah, Dr,
Azita Emami and Dr. Reza Mohammadi as well as Mrs. Akram
Shahrokhi, Mr. Kazem Nazmdeh and Mr. Mohammad Reza
Soofinejad for their support.
References
Abbema, E. A., Van Assema, P., Kok, G. J., De Leeuw, E., & De Vries, N. K. (2004). Effect
evaluation of a comprehensive community intervention aimed at reducing
socioeconomic health inequalities in the Netherlands? Health Promotion
International, 19(2), 141–156.
Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological
Bulletin, 107, 238–246.
Boomsma A. (1982). Robustness of LISREL against small sample sizes in factor
analysis models. In: Joreskog KG, Wold H, editors. Systems under indirection
observation: Causality, structure, prediction (Part I) Amsterdam, Netherlands:
North Holland, 149–173.
Bostan, C., Oberhauser, C., Stucki, G., Bickenbach, J., & Cieza, A. (2015). Which
environmental factors are associated with lived health when controlling for
biological health?—a multilevel analysis. BMC Public Health, 27(15), 508.
Braveman, P. (2011). Accumulating knowledge on the social determinants of health
and infectious disease. Public Health Reports, 126(Suppl. 3), 28–230.
 Z. Mosallanezhad et al. / Archives of Gerontology and Geriatrics 70 (2017) 123–129
Brodersen, N. H., Steptoe, A., Boniface, D. R., & Wardle, J. (2007). Trends in physical
activity and sedentary behaviour in adolescence: Ethnic and socioeconomic
differences. British Journal of Sports Medicine, 41, 140–144.
Brown, D. W., Balluz, L. S., Heath, G. W., Moriarty, D. G., Ford, E. S., Giles, W. H., et al.
(2003). Associations between recommended levels of physical activity and
health-related quality of life Findings from the 2001 Behavioral Risk Factor
Surveillance System (BRFSS) survey? Preventive Medicine, 37(5), 520–528.
Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A.
Bollen, & J. S. Long (Eds.), Testing structural equation models (pp. 136–162).
Newbury Park, CA: Sage.
Chakravarty, E. F., Hubert, H. B., Krishnan, E., Bruce, B. B., Lingala, V. B., & Fries, J. F.
(2012). Lifestyle risk factors predict disability and death in healthy aging adults?
American Journal of Medicine, 125(2), 190–197.
Cockerham, W. C. (2005). Health lifestyle theory and the convergence of agency and
structure. Journal of Health and Social Behavior, 46, 51.
Davin, B., Paraponaris, A., & Verger, P. (2005). Demographic and socioeconomic
factors associated with needs for home assistance among community-dwelling
elderly: A study from the French Home Survey Handicaps-Disabilities-
Dependence. Rev Epidemiol Sante Publique, 53(5), 509–524 [Article in French].
Dowd, J. B., & Goldman, N. (2006). Do biomarkers of stress mediate the relation
between socioeconomic status and health? Journal of Epidemiology and
Community Health, 60(7), 633–639.
Dragano, N., Bobak, M., Wege, N., Peasey, A., Verde, P. E., Kubinova, R., et al. (2007).
Neighbourhood socioeconomic status and cardiovascular risk factors: A
multilevel analysis of nine cities in the Czech Republic and Germany. BMC Public
Health, 7, 255.
Elovainio, M., Kivimäki, M., Kortteinen, M., & Tuomikoski, H. (2001). Socioeconomic
status: Hostility and health. Personality and Individual Differences, 31(3), 303–
315.
Eriksson, B. G., Mellström, D., & Svanborg, A. (1987). A medical-social intervention in
a 70-year-old Swedish population: A general presentation of methodological
experience. Comprehensive Gerontology. Section C, 1, 49–56.
Fiatarone Singh, M. A. (2002). Exercise comes of age: Rationale and
recommendations for a geriatric exercise prescription. Journal of Gerontology:
Medical Sciences, 57A(5), M262–M282.
Frändin, K., & Grimby, G. (1994). Assessment of physical activity, fitness, and
performance in 76-year-olds. Scandinavian Journal of Medicine & Science in
Sports, 4, 41–46.
Gagné, P., & Hancock, G. R. (2006). Measurement model quality, sample size, and
solution propriety in confirmatory factor models. Multivariate Behavioral
Research, 41, 65–83.
Hanson, M. D., & Chen, E. (2007). Socioeconomic status and health behaviors in
adolescence: A review of the literature. Journal of Behavioral Medicine, 30(3),
263–285 [Epub 2007 May 20].
Hayduk, L. A. (1996). LISREL issues, debates, and strategies. Baltimore and London:
The Johns Hopkins University Press.
Hillsdon, M., Lawlor, D. A., Ebrahim, S., & Morris, J. N. (2008). Physical activity in
older women: Associations with area deprivation and with socioeconomic
position over the life course: Observations in the British Women’s Heart and
Health Study. Journal of Epidemiology and Community Health, 62, 344–350.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure
analysis: Conventional criteria versus new alternatives. Structural Equation
Modeling, 6, 1–55.
Iversen, A. C., & Kraft, P. (2006). Does socio-economic status and health
consciousness influence how women respond to health related messages in
media? Health Education Research, 21(5), 601–610.
Kıvanç, S. A., Akova-Budak, B., Olcaysü, O. O., & Çevik, S. G. (2016). Sociodemographic
status of severely disabled and visually impaired elderly people in Turkey.
Arquivos Brasileiros de Oftalmologia, 79(1), 24–29.
Kim, H., Lee, T., Lee, S., Kim, K., Lee, S., Kam, S., et al. (2012). Factors associated with
ADL and IADL dependency among Korean centenarians: Reaching the 100-year-
old life transition? International Journal of Aging and Human Development, 74(3),
243–264.
Lantz, P. M., Lynch, J. W., House, J. S., Lepkowski, J. M., Mero, R. P., Musick, M. A., et al.
(2001). Socioeconomic disparities in health change in a longitudinal study of US
adults: The role of health-risk behaviors. Social Science & Medicine, 53(1), 29–40.
Lee, S. G., & Jeon, S. Y. (2005). The relations of socioeconomic status to health status:
Health behaviors in the elderly. Journal of Preventive Medicine & Public Health, 38
(2), 154–162.
129
Maas, J., Verheij, R. A., Spreeuwenberg, P., & Groenewegen, P. P. (2008). Physical
activity as a possible mechanism behind the relationship between green space
and health: A multilevel analysis. BMC Public Health, 8, 206–219.
Mackenbach, J. P., & Bakker, M. J. (2003). European network on interventions and
policies to reduce inequalities in health. Tackling socioeconomic inequalities in
health: Analysis of European experiences. Lancet, 362(9393), 1409–1414.
Mosallanezhad, Z., Salavati, M., Hellström, K., Sotoudeh, G. R., Nilsson Wikmar, L., &
Frändin, K. (2011). Cross-cultural adaptation, reliability and validity of the
Persian version of the modified falls efficacy scale. Disability and Rehabilitation,
4, 4.
Mosallanezhad, Z., Hörder, H., Salavati, M., Nilsson Wikmar, L., & Frändin, K. (2012).
Health, physical activity and physical functioning in Swedish and Iranian 75-
year-olds—a comparison. Archives of Gerontology and Geriatrics02.007.
Mulder, B. C., de Bruin, M., Schreurs, H., van Ameijden, E. J., & van Woerkum, C. M.
(2011). Stressors and resources mediate the association of socioeconomic
position with health behaviours. BMC Public Health, 13(11), 798.
Muthén, L. K., & Muthén, B. O. (2007). Mplus user’s guide, 5th ed. Los Angeles, CA:
Muthén n.
Park, E. J., Sohn, H. S., Lee, E. K., & Kwon, J. W. (2014). Living arrangements, chronic
diseases, and prescription drug expenditures among Korean elderly:
Vulnerability to potential medication underuse. BMC Public Health, 16(14), 1284.
Pruchno, R. A., Wilson-Genderson, M., & Cartwright, F. P. (2012). The texture of
neighborhoods and disability among older adults. Journals of Gerontology Series
B: Psychological Sciences and Social Sciences, 67(1), 89–98.
Rodriguez, C. J., Sciacca, R. R., Diez-Roux, A. V., Boden-Albala, B., Sacco, R. L., Homma,
S., et al. (2004). Relation between socioeconomic status, race-ethnicity, and left
ventricular mass. Hypertension, 43, 775–779.
Rutt, C. D., & Karen, C. J. (2005). Examining the relationships among built
environment physical activity, and body mass index in El Paso, TX. Preventive
Medicine, 40(6), 831–841.
Schöllgen, I., Huxhold, O., Schüz, B., & Tesch-Römer, C. (2011). Resources for health:
Differential effects of optimistic self-beliefs and social support according to
socioeconomic status? Health Psychology, 30(3), 326–335.
Schulz, A. J., House, J. S., Israel, B. A., Mentz, G., Dvonch, J. T., Miranda, P. Y., et al.
(2008). Relational pathways between socioeconomic position and
cardiovascular risk in a multiethnic urban sample: Complexities and their
implications for improving health in economically disadvantaged populations.
Journal of Epidemiology and Community Health, 62, 638–646.
Sonn, U., Törnquist, K., & Svensson, E. (1999). The ADL taxonomy-from individual
categorical data to ordinal categorical data? Scandinavian Journal of Occupational
Therapy, 6(1), 11–20.
Stewart, K. J. (2005). Physical activity and aging. Annals of the New York Academy of
Sciences, 1055, 193–206.
Sun, W., Watanabe, M., Tanimoto, Y., Shibutani, T., Kono, R., Saito, M., et al. (2007).
Factors associated with good self-rated health of non-disabled elderly living
alone in Japan: A cross-sectional study. BMC Public Health, 7, 297.
Teymoori, F., & Foroozan, M. (2005). Validity and reliability assessment of the Persian
version of the Leipad quality of life questionnaire for elderly people. The University
of Social Welfare and Rehabilitation Sciences.
Tse, W. S., Rochelle, T. L., & Cheung, J. C. (2011). The relationship between
personality, social functioning, and depression: A structural equation modeling
analysis. International Journal of Psychology, 46(June (3)), 234–240.
Umberson, D., & Montez, J. K. (2010). Social relationships and health: A flashpoint
for health policy. Journal of Health and Social Behavior, 51(Suppl), S54–S66.
http://dx.doi.org/10.1177/0022146510383501.
Velicer, W. F., & Fava, J. L. (1998). Effects of variable and subject sampling on factor
pattern recovery. Psychological Methods, 3, 231–251.
Watt, H. C., Carson, C., Lawlor, D. A., Patel, R., & Ebrahim, S. (2009). Influence of life
course socioeconomic position on older women’s health behaviors: Findings
from the British Women’s Heart and Health Study. American Journal of Public
Health, 99(2), 320–327.
Wormald, H., Waters, H., Sleap, M., & Ingle, L. (2006). Participants’ perceptions of a
lifestyle approach to promoting physical activity: Targeting deprived
communities in Kingston-upon-Hull. BMC Public Health, 4(6), 202.
Yunhwan, L. (2000). The predictive value of self-assessed general, physical, and
mental health on functional decline and mortality in older adults. Journal of
Epidemiology and Community Health, 54, 123–129.