Journal of Physical Activity and Health 2006, 3, 90-101

© 2006 Human Kinetics, Inc.

Socioeconomic Predictors of a
Sedentary Lifestyle: Results From
the 2001 National Health Survey
Adrienne Brown and Mohammad Siahpush

Background: Regular physical activity reduces the risk of a number of diseases,

prevents obesity, and has positive psychological effects. Approximately one-third
of the Australian population has been reported as totally sedentary. We investigated
socioeconomic predictors of being sedentary in a nationally representative sample
of Australian adults. Methods: We analyzed data from 8643 females and 7600
males who responded to the 2001 National Health Survey. Multivariate logistic
regression was used to examine the association of being sedentary with a range of
socioeconomic measures. Results: Adjusting for demographics, body-mass index,
and smoking, we found that low socioeconomic status, indicated by low education
level, blue-collar occupation, low income and area social disadvantage, increased
the probability that people were sedentary. Conclusions: This research highlights
that targeting people from lower socioeconomic backgrounds with strategies to
increase participation in physical activity may reduce morbidity and mortality
associated with being sedentary.

Key Words: sedentary, socioeconomic predictors, Australia

Regular physical activity, combined with appropriate diet, is one of the most effec-
tive means of controlling obesity and maintaining ideal body weight,1, 2 reduces
the risk of cardiovascular disease and diabetes, colon cancer, breast cancer,3 and
osteoporosis,1 and has positive psychological and social effects.1 However, despite
a high level of knowledge regarding the benefits of physical activity, its practice is
declining and the proportion of physically inactive people in Australia is increas-
ing.4 Approximately one-third of the Australian population has been reported as
totally sedentary, reporting no activity in the 2 wk prior to survey.5-7 Given that lack
of physical activity is responsible for about 7% of the total burden of disease and
disability in Australia8 and is one of the 10 leading causes of death and disability in
the world,3 a fundamental goal of public health initiatives should be the promotion
of physical activity among those who are sedentary.

Brown is with the ORYGEN Research Centre, The University of Melbourne, Parkville, Victoria 3052,
Australia. Siahpush is with the Centre for Behavioural Research in Cancer, The Cancer Council Victoria,
Carlton, Victoria 3053, Australia.

90
 Socioeconomic Predictors of a Sedentary Lifestyle 91

Socioeconomic status (SES) has emerged as an important contributor to social

inequalities in health. Adults from lower socioeconomic backgrounds have been
reported as more likely to perceive health-related behaviors as compromising qual-
ity of life, to be least knowledgeable about health-related recommendations and
guidelines, and less likely to perceive current behavior as health compromising.9 In
particular, SES has been correlated with participation in physical activity; people
living in lower SES neighborhoods have been found to engage in less physical
activity.10 One reason provided for this is reduced availability and accessibility of
physical activity resources.11 Research using education as a measure of SES has
found strong associations between participation in physical activity and higher
socioeconomic position.9, 12, 13 Engagement in physical activity decreases from
high to low levels of education,6, 14 and people who have not completed secondary
school or who are less well educated are significantly more likely to be inactive,
or sedentary.7, 15 The association between occupation and participation in physical
activity is not clear. Salmon, Owen, Bauman et al. found that professional workers
were more likely to engage in leisure-time physical activity.6 Homemakers, skilled,
and unskilled workers were less likely to engage in vigorous leisure-time physi-
cal activity necessary for health or fitness benefits, and less-skilled workers were
the least likely to report any form of leisure-time physical activity, compared to
professionals.6 However, other research has found that unskilled workers engage in
significantly more occupational and total physical activity than workers in higher
status occupations.16 Higher income has been associated with an increase in physi-
cal activity5 and people with low incomes have been found to be more likely to be
sedentary.7 However, income as a measure of SES has received scant attention.
Past research has typically used specific age groups12, 17 and restricted popula-
tions15 to assess sociodemographic and behavioral correlates of participation in
various types of physical activity; walking, moderate, high-level, occupational, or
household physical activity.14, 18 Few studies have investigated predictors of being
sedentary, which is considered a more robust measure than expenditure measures
of physical activity. It has also been argued that the greatest benefit will result from
the sedentary becoming active.7 Furthermore, few studies have used a wide range
of SES measures in determining the magnitude and statistical significance of the
independent effect of each as a predictor of sedentary behavior. Since approximately
30% of the Australian population are sedentary,5, 6 analysis of factors that prevent
people from being active appears essential.
We investigated predictors of being sedentary in a nationally representative
sample of Australian adults using a wide range of socioeconomic measures. Adjust-
ing for demographic factors, body-mass index (BMI), and smoking, which have
been associated with physical activity participation,13, 15, 19 we hypothesized that
lower socioeconomic position would be a barrier towards activity.

Methods
Sample Selection
Data were obtained from the 2001 National Health Survey (NHS), which was con-
ducted by the Australian Bureau of Statistics (ABS) using a stratified multistage
92 Brown and Siahpush

area sample of private dwellings.20 Trained ABS interviewers performed face-to-

face interviews with one selected resident from each dwelling. The response rate
was 92% following sample loss, with a total of 26,863 respondents. All aspects
of the NHS’s implementation were designed to conform to Information Privacy
Principles set out in the Privacy Act 1988, and the Privacy Commissioner was
informed of the details of the proposed surveys. The ABS sought cooperation of
households in the survey and the confidentiality of all information provided by
respondents was guaranteed.20

Measurement
The ABS measured physical activity undertaken for recreation, sport, or health/
fitness purposes during the 2 wk prior to interview. Respondents were asked whether,
during the previous 2 wk, they did any walking for sport, recreation, or fitness, any
moderate exercise (defined as exercise that caused a moderate increase in the heart
rate or breathing of the respondent), or any vigorous exercise (defined as exercise
that caused a large increase in the respondent’s heart rate or breathing). Interpreta-
tion of these definitions reflected the respondent’s perception of moderate or vigor-
ous exercise or walking, and the purpose of that activity. As such, variation was
expected in responses according to the type of activity, the intensity with which it
was performed, the level of fitness of the subject, and their general health. Physical
activity undertaken in the course of work and activity undertaken for reasons other
than recreation, sport, or health/fitness were not recorded.
Respondents were asked the number of times they had engaged in each type
of exercise (walking, moderate, or vigorous) and the total amount of time spent
(hours and minutes) doing that exercise, in the previous 2 wk. Using information
about frequency, duration, and intensity of exercise, the ABS derived an exercise
level for each respondent to indicate the quality of the activities undertaken in terms
of maintaining heart, lung, and muscle fitness. They used the following formula:
number of times activity was undertaken × average time × intensity per session.
Intensity was expressed as a multiple of the resting metabolic rate and was estimated
for each category of exercise. The intensity value of walking was 3.5, of moderate
exercise 5.0, and of vigorous exercise 7.5. A score was derived for each of the three
categories of exercise and then summed to provide a total for each respondent for
that 2 wk period. The ABS then classified respondents as being sedentary (scores
< 100), as engaging in low-level exercise (scores < 100 to 1600), moderate level
exercise (scores 1600 to 3200 or > 3200 and < 2 h vigorous exercise), or as engag-
ing in a high level of exercise (scores > 3200 and ≥ 2 h vigorous exercise). These
categories were used in the bivariate analyses (Tables 1 and 2).
The outcome in the multivariate logistic regression analysis (Table 3) was
sedentary versus active. Subjects were considered sedentary if they engaged in no
physical activity for the purpose of recreation, sport, or health/fitness and active if
they engaged in physical activity, regardless of intensity.
The following demographic and health behavior characteristics were used as
covariates: age, country of birth, marital status, region of residence, body-mass
index (BMI), and smoking status. Marital status was categorized into married (living
with a partner in a spouse/de facto relationship) and not married (not living with a
partner). BMI, a measure of a person’s weight in relation to their height (kg/m2),
 Socioeconomic Predictors of a Sedentary Lifestyle 93

was used to classify respondents into weight categories. A BMI < 25 was defined
as not overweight, BMI 25 to < 30 kg/m2 as overweight, and BMI ≥ 30 kg/m2 as
obese. 21 BMI was calculated using self-reported height and weight.
Education level, occupation, household income, and index of relative socio-
economic disadvantage (IRSD) were investigated as socioeconomic predictors of
being sedentary. Education was categorized into the following five groups: tertiary
degree/diploma; vocational qualification; unspecified post-school qualification;
completed secondary school; and not completed secondary school. Occupation
was categorized into the following five groups: professional; white-collar; blue-
collar; unemployed; and not in the labor force. Respondent’s descriptions of their
occupation in their main job or business and the main tasks or duties performed
were coded using the Australian Standard Classification of Occupations (ASCO).
Professionals included managers, administrators, professionals, and associate
professionals; white-collar workers included clerical, sales, and service workers;
and blue-collar workers included trades persons and related workers, intermediate
production and transport workers, and laborers and related workers. Unemployed
persons were those who were not employed in the reference week, but who were
looking for work or waiting to start work in the previous or following 4 wk. Persons
who were not employed or unemployed were classified as not in the labor force.
The IRSD is an area socioeconomic index compiled at the collection district level
by the ABS and includes such variables as the income, education, occupation,
housing, household composition, and English fluency of residents.20, 22 IRSD was
classified into quintiles, as was household income.

Statistical Analysis
Cases for which a minimum (< 1.0%) of responses were coded as “not stated” or
“not applicable” were excluded from analyses. Females and males were analyzed
separately because of an interaction effect of sex with age, occupation, and smoking
status. Multivariate logistic regression was used to investigate associations between
being sedentary and each of the covariates. Due to the multi-stage nature of the
sampling design, standard errors in the logistic regression analyses needed to be
adjusted. Such adjustment requires information on the primary sampling units for
each observation in the dataset. Since the ABS did not provide this information, we
used bootstrapping to estimate standard errors. The bootstrap is a data-dependent
method of computing the standard error and is suitable for situations where no
formula is available to compute standard errors or available formulas make unten-
able assumptions.23 Analyses were conducted using Stata 8.2.24

Results
The sample comprised 53.2% (n = 8643) females and 46.8% (n = 7600) males age
18 y and over. Of the total sample, 18.2% were age 18 to 29 y, 33.4% were age 30
to 44 y, 24.5% were age 45 to 59 y, and 23.9% were age 60 y and over. Of females,
29.3% were sedentary, 43.1% had engaged in a low level of physical activity, 23.5%
had engaged in a moderate level, and 4.0% had engaged in a high level of physical
activity in the 2 wk before the survey. Of males, 28.5% were sedentary, 35.1% had
engaged in a low level of physical activity, 27.6% had engaged in a moderate level,
94 Brown and Siahpush

Table 1 Sample Characteristics and Bivariate Associations

Between Covariates and Physical Activity Level (sedentary, low,
medium, high) Among Females (n = 8643)
Physical activity level
Sedentary Low Moderate High
% in (n = 2534) (n = 3725) (n = 2035) (n = 349)
sample % % % % χ2
Age 230.68**
18-29 18.2 23.2 43.8 25.1 7.9
30-44 33.3 26.6 46.7 21.8 4.8
45-59 23.4 28.5 43.2 25.5 2.9
60+ 25.1 38.2 37.7 22.8 1.2
Country of birth 39.31**
Australia 73.4 27.5 44.6 23.8 4.1
Other 26.6 34.3 39.0 22.9 3.9
Marital status 18.77**
Not married 54.2 29.3 44.5 22.8 3.4
Married 45.8 29.3 41.4 24.5 4.8
Region of residence 13.78*
Metropolitan 66.8 28.6 43.4 23.6 4.4
Inner regional 20.8 29.5 43.3 23.6 3.6
Other 12.4 32.6 41.4 23.3 2.7
Education 332.38**
Tertiary 26.0 18.4 47.3 28.6 5.7
Vocational 19.2 26.6 44.7 23.7 5.0
Post-school qual not stated 3.4 26.3 44.0 27.0 2.7
Completed secondary 11.6 28.0 42.2 23.5 6.4
Less than secondary 39.7 38.5 39.7 19.9 1.9
Occupation 250.82**
Professional 22.0 18.4 48.3 26.7 6.6
White-collar 24.8 27.0 45.0 23.2 4.8
Blue-collar 6.9 37.1 38.9 20.9 3.2
Unemployed 2.8 20.7 47.1 26.9 5.4
Not in labor force 43.5 35.5 39.8 22.4 2.4
Household incomea 221.66**
First quintile 15.6 18.0 46.6 29.3 6.2
Second quintile 15.1 27.1 46.6 21.7 4.7
Third quintile 13.5 26.6 48.2 22.2 2.9
Fourth quintile 17.9 33.5 42.5 21.9 2.1
Fifth quintile 19.7 37.5 37.9 21.9 2.7
Unknown 18.3 30.0 39.7 24.4 5.8
ISRDb 107.11**
First quintile 22.5 23.4 43.0 28.6 5.0
Second quintile 22.0 28.2 44.5 22.7 4.6
Third quintile 18.7 29.6 44.4 22.0 4.0
Fourth quintile 18.6 30.0 45.2 21.5 3.3
Fifth quintile 18.3 36.8 38.1 22.1 3.0
Body-mass index (BMI) 107.14**
Not overweight 57.8 27.7 41.5 25.4 5.4
Overweight 25.6 29.6 44.9 22.8 2.7
Obese 16.6 34.4 45.9 18.4 1.3
Smoking status 23.74**
Non/ex-smoker 78.0 28.0 43.9 24.0 4.1
Current smoker 22.0 33.8 40.4 22.0 3.8
Note. Source: National Health Survey (NHS) 2001. a First quintile denotes high income and fifth quintile denotes
low income; b First quintile denotes low disadvantage and fifth quintile denotes high disadvantage; χ2 tests: *P <
0.05; **P < 0.01.
 Socioeconomic Predictors of a Sedentary Lifestyle 95

Table 2 Sample Characteristics and Bivariate Associations

Between Covariates and Physical Activity Level (sedentary, low,
medium, high) Among Males (n = 8643)
Physical activity level
Sedentary Low Moderate High
% in (n = 2165) (n = 2669) (n = 2094) (n = 672)
sample % % % % χ2
Age 356.52**
18-29 18.3 19.5 33.0 28.1 19.4
30-44 33.6 28.6 36.4 25.4 9.5
45-59 25.7 30.2 37.0 26.5 6.4
60+ 22.5 33.7 32.7 31.4 2.1
Country of birth 25.33**
Australia 71.5 27.3 35.5 27.5 9.7
Other 28.5 31.4 34.2 27.7 6.7
Marital status 67.42**
Not married 59.2 29.6 37.0 26.5 6.9
Married 40.8 26.9 32.3 29.1 11.7
Region of residence 24.62**
Metropolitan 66.5 27.1 36.0 27.5 9.4
Inner regional 19.6 29.8 34.5 27.1 8.5
Other 13.9 33.2 31.8 28.3 6.7
Education 309.05**
Tertiary 26.8 18.3 37.4 31.8 12.4
Vocational 30.6 30.0 35.7 26.7 7.6
Post-school qual not stated 1.8 23.0 45.9 23.7 7.4
Completed secondary 11.5 25.0 31.2 28.6 15.2
Less than secondary 29.4 37.8 33.3 24.4 4.5
Occupation 278.18**
Professional 31.6 21.5 38.2 28.3 12.0
White-collar 9.8 23.6 36.5 26.9 13.0
Blue-collar 28.2 36.5 33.1 21.7 8.7
Unemployed 4.0 19.8 33.1 37.7 9.4
Not in labor force 26.2 31.5 33.3 31.6 3.6
Household incomea 156.76**
First quintile 22.2 21.0 36.0 30.5 12.5
Second quintile 16.7 27.7 37.2 26.4 8.7
Third quintile 15.1 30.0 35.8 25.9 8.3
Fourth quintile 13.9 33.3 34.4 27.6 4.7
Fifth quintile 15.2 34.7 33.2 27.6 4.5
Unknown 16.8 28.3 33.6 26.2 11.9
ISRDb 128.58**
First quintile 22.6 20.5 37.6 30.5 11.3
Second quintile 22.4 26.4 38.0 25.6 10.0
Third quintile 18.7 30.2 32.7 28.9 8.2
Fourth quintile 19.4 31.7 34.4 26.5 7.4
Fifth quintile 16.9 36.3 31.4 25.9 6.3
Body-mass index (BMI) 68.55**
Not overweight 42.1 26.9 33.2 29.6 10.3
Overweight 42.1 28.0 35.7 27.6 8.7
Obese 15.8 34.2 38.6 22.0 5.2
Smoking status 104.94**
Non/ex-smoker 72.6 25.3 36.2 29.1 9.5 72.6
Current smoker 27.4 37.0 32.4 23.6 7.1 27.4
Note. Source: National Health Survey (NHS) 2001. a First quintile denotes high income and fifth quintile denotes low
income; b First quintile denotes low disadvantage and fifth quintile denotes high disadvantage; χ2 tests: **P < 0.01.
96 Brown and Siahpush

Table 3 Odds Ratios (OR) and 95% Confidence Intervals (CI)

from Logistic Regression of the Likelihood of Being Sedentary on
Socio-Demographic Variables and BMI Among Australians Over
18 y of Age (n = 16,243)
Females Males
OR (CI) P a
OR (CI) Pa
Age P < 0.01 P < 0.01
18-29 1.0 1.0
30-44 1.14 (0.98-1.32) 1.63 (1.36-1.94)
45-59 1.15 (0.97-1.36) 1.88 (1.56-2.28)
60+ 1.52 (1.26-1.83) 2.43 (1.95-3.04)
Country of birth P < 0.01 P < 0.01
Australia 1.0 1.0
Other 1.38 (1.24-1.54) 1.19 (1.06-1.33)
Marital status P < 0.01 P < 0.05
Not married 1.0 1.0
Married 0.87 (0.78-0.97) 0.87 (0.78-0.98)
Region of residence P = 0.47 P = 0.64
Metropolitan 1.0 1.0
Inner regional 0.95 (0.84-1.08) 0.98 (0.85-1.13)
Other 1.06 (0.91-1.22) 1.06 (0.91-1.24)
Education P < 0.01 P < 0.01
Tertiary 1.0 1.0
Vocational 1.32 (1.13-1.56) 1.37 (1.15-1.62)
Post-school qual not stated 1.31 (0.98-1.75) 0.99 (0.62-1.59)
Completed secondary 1.48 (1.21-180) 1.44 (1.18-1.77)
Less than secondary 1.94 (1.68-2.24) 1.84 (1.55-2.18)
Occupation P < 0.01 P < 0.01
Professional 1.0 1.0
White-collar 1.22 (1.04-1.43) 0.95 (0.76-1.19)
Blue-collar 1.60 (1.29-2.00) 1.52 (1.30-1.77)
Unemployed 0.72 (0.50-1.02) 0.50 (0.36-0.70)
Not in labor force 1.25 (1.05-1.48) 0.79 (0.65-0.97)
Household incomeb P < 0.01 P < 0.01
First quintile 1.0 1.0
Second quintile 1.43 (1.18-1.74) 1.15 (0.95-1.38)
Third quintile 1.19 (0.97-1.46) 1.16 (0.96-1.41)
Fourth quintile 1.42 (1.15-1.74) 1.33 (1.08-1.63)
Fifth quintile 1.63 (1.32-2.01) 1.60 (1.29-1.99)
Unknown 1.51 (1.25-1.83) 1.36 (1.13-1.63)
ISRDc P < 0.05 P < 0.01
First quintile 1.0 1.0
Second quintile 1.13 (0.97-1.31) 1.13 (0.96-1.34)
Third quintile 1.13 (0.96-1.33) 1.28 (1.07-1.53)
Fourth quintile 1.05 (0.89-1.24) 1.33 (1.11-1.58)
Fifth quintile 1.30 (1.11-1.53) 1.49 (1.24-1.79)
Body-mass index (BMI) P < 0.05 P < 0.01
Not overweight 1.0 1.0
Overweight 1.00 (0.89-1.13) 1.00 (0.89-1.12)
Obese 1.20 (1.05-1.36) 1.28 (1.10-1.49)
Smoking status P < 0.01 P < 0.01
Non/ex-smoker 1.0 1.0
Current smoker 1.36 (1.21-1.52) 1.72 (1.53-1.93)
Note. Source: National Health Survey (NHS) 2001. aP -values pertain to Wald tests for the effects of the set of dummy
variables representing each covariate; bFirst quintile denotes high income and fifth quintile denotes low income; cFirst
quintile denotes low disadvantage and fifth quintile denotes high disadvantage.
 Socioeconomic Predictors of a Sedentary Lifestyle 97

and 8.8% had engaged in a high level of physical activity in the 2 wk before the
survey. Sample characteristics and the distribution of physical activity levels for
each covariate are shown in Table 1 for females and Table 2 for males. Chi-square
tests revealed significant bivariate associations between level of physical activity
and all covariates, for females and males.
Significant associations were found between all demographic and behavioral
factors and being sedentary, apart from region of residence. While the odds of being
sedentary increased for each successive age category for both genders, age was
a stronger predictor of being sedentary for males. The results suggest that people
who were born in countries other than Australia and who were “obese” were more
likely to be sedentary and those living with a partner were less likely to be sedentary.
“Overweight” females and males were no more likely to be sedentary than those
who were “not overweight.” While the odds of being sedentary were 36% greater
for females who were current smokers, they were 72% greater for males who were
current smokers, compared to non/ex-smokers.
Adjusting for demographic and behavioral factors, we found significant asso-
ciations between all SES measures and being sedentary for females and males (see
Table 3). A strong inverse relationship was found between level of education and
being sedentary for both genders, with the exception of those who did not specify
their post-school qualification. The results generally suggest that female and male
blue-collar workers were most likely to be sedentary, compared to those working
in other occupations. The probability of being sedentary was greater for females
working in white-collar occupations, compared to professionals. The probability
of being sedentary was smaller for unemployed than employed people; however,
women who were not in the labor force were more likely to be sedentary whereas
men not in the labor force were less likely to be sedentary, compared to profes-
sionals.
For males, the odds of being sedentary increased with each successive income
category, compared to the highest income category. The results generally suggest
that females with low incomes had a greater probability of being sedentary. Simi-
larly, the odds of being sedentary increased with each successive quintile of area
social disadvantage for males. For females, the results generally suggest that those
living in more disadvantaged areas were more likely to be sedentary.

Discussion
Using a wide range of socioeconomic measures, we investigated predictors of being
sedentary in a nationally representative sample of Australian adults. As predicted,
lower socioeconomic position was found to be a barrier towards physical activity.
Low education level, blue-collar occupation, low income, and area social disad-
vantage were all significant predictors of sedentary behavior.
The strong inverse relationship between education level and being sedentary
for both genders, which was also found in research examining physical activity
level,12-14 provides further support for the necessity of promoting the benefits of
being active in those with low educational attainment. While household income
and area social disadvantage were significantly associated with being sedentary
for both genders, a strong inverse relationship existed between these measures of
SES and being sedentary for males. The presence of a relationship between being
98 Brown and Siahpush

sedentary and area social disadvantage indicates that neighborhood SES has a con-
textual effect on being sedentary over that of individual measures. This suggests that
social, physical, cultural, or economic factors may predispose a sedentary lifestyle.
Strategies to increase activity should therefore be beneficial if they target persons at
the “community-level.” Well-maintained, continuous, and well-lit sidewalks, safe
and convenient crossings, and pedestrian and bicycle-friendly neighborhoods are
examples of community-level initiatives that have been found to increase participa-
tion.25 Increased membership to local sporting groups and organizations (i.e., golf,
football, surfing, tennis clubs) is another strategy that would enhance health and
fitness, along with providing benefits associated with social capital such as lower
delinquency and crime rates26 and improved general health.27
Working in a blue-collar occupation was the strongest occupational predictor
of being sedentary for both genders. This finding may represent the increased risk
of cardiovascular disease found in less skilled workers6 and implies that they should
be targeted as a group. One suggestion would be to gain support from companies
that employ blue-collar workers. The advent of sporting groups and clubs within
such organizations would not only benefit the individual but also increase produc-
tivity within the organization.28 However, as physical activity for purposes other
than recreation, sport, or health/fitness purposes was not recorded, it is possible
that those classified as sedentary in the current study may have been engaging
in occupation- or home-related physical activity, such as housework or garden-
ing. While working in a white-collar occupation and not being in the labor force
reduced the likelihood that males were sedentary, the opposite effect was found
for females. This finding may represent increased responsibilities of females with
respect to motherhood and associated lack of time, which are proven barriers to
physical activity.29, 30 Thus women, especially mothers, require support to integrate
physical activity into their daily routines. The finding that unemployment reduced
the risk of being sedentary again suggests that time is an important determinant of
participation in physical activity. Future research that acknowledges leisure-time
as well as home- and occupation-related physical activity is required to further
clarify these findings.
Age and country of birth had the same effect on sedentary behavior as was
found in past research examining engagement in physical activity.12, 14 However,
increasing age was a much stronger predictor of being sedentary for males and being
born in a country other than Australia was a stronger predictor of being sedentary
for females. These findings suggest that health promotion activities promoting
ongoing and increased participation in physical activity should target older people
and those born outside of Australia. People living in a marriage-like relationship
were significantly less likely to lead sedentary lifestyles, a finding inconsistent with
past research.6, 29 This result may be accounted for by differing definitions of mar-
riage; while past researchers used traditional definitions of marriage, we considered
people living with a partner as married. Diverging from past research, which has
identified rural and urban women as having a different pattern of determinants
of leisure-time physical activity,30 region of residence did not predict sedentary
behavior. The finding that people who were “obese” were significantly more likely
to be sedentary is very concerning given the proven independent effects of obesity31
and physical inactivity8 on health. The finding that smokers were more likely to be
sedentary is similarly concerning given that both smoking and sedentary behaviors
 Socioeconomic Predictors of a Sedentary Lifestyle 99

produce serious consequences for health. Initiatives that address the health effects
of smoking could therefore benefit from increasing smokers’ awareness of the
benefits of being active, over and above weight control. Physical activity is also a
valuable distraction technique when quitting.
Some limitations of the NHS data are worth mentioning. First, since physical
activity undertaken for reasons other than recreation, sport, or health/fitness was
not examined, levels reported may not represent overall levels of activity. Second,
due to the nature of the ABS data we were unable to use multi-level modeling to
determine the influence of individual versus area-level measures of SES. Third, the
cross-sectional nature of the survey meant that while associations could be identi-
fied it was not possible to infer causal relationships. Fourth, while we identified
important associations between SES and sedentary behavior, further research is
required to explore barriers to participation in these groups.

Conclusions
The current research indicates that all indicators of low SES are powerful individual
contributors to being sedentary. Priority should be given to people with low levels
of education and low incomes, who work in blue-collar occupations and reside in
more disadvantaged areas. Strategies that promote physical activity through mem-
bership or involvement in local sporting groups, through encouraging organizations
to support physical activity in their blue-collar workers, and through teaching people
how to manage their time to allow regular physical activity, could help those who
are sedentary become more active. The promotion of regular walking for physi-
cal activity is the most achievable first step as it is convenient, free, and has been
found to result in weight loss.32 However, future public health strategies must also
aim to increase the proportion of people engaging in other types of moderate- and
vigorous-intensity physical activity, in addition to walking, which may be enjoyable
and is necessary to achieve additional health benefits.33

Acknowledgments
This research was supported by a grant from the Victorian Health Promotion Founda-
tion (VicHealth).

References
1. World Health Organization. Readings on physical activity: Regional Office for the
Eastern Mediterranean, 2002.
2. Ball K, Owen N, Salmon J et al. Associations of physical activity with body weight
and fat in men and women. Int J Obes. 2001;25:914-919.
3. World Health Organization. The World Health Report. Reducing risks, promoting
healthy life. Geneva: World Health Organization, 2002:1-20.
4. Armstrong T, Bauman A, Davies J. Physical activity patterns of Australian adults:
Results of the 1999 National Physical Activity Survey. Canberra: Australian Institute
of Health and Welfare, 2000.
5. Bauman A, Owen N, Rushworth L. Recent trends and socio-demographic determinants
of exercise participation in Australia. Community Health Studies. 1990;14:19-26.
100 Brown and Siahpush

6. Salmon BA, Owen N, Bauman A et al. Leisure-time, occupational, and household

physical activity among professional, skilled, and less-skilled workers and homemak-
ers. Prev Med. 2000;30:191-199.
7. Owen N, Bauman A. The descriptive epidemiology of a sedentary lifestyle in adult
Australians. Int J Epidemiol. 1992;21:305-310.
8. Mathers C, Voc T, Stevenson C. The burden of disease and injury in Australia. Canberra:
Australian Institute of Health and Welfare, 1999.
9. Burton NW, Turrell G, Oldenburg B. Participation in recreational physical activity:
why do socioeconomic groups differ? Health Educ Behav. 2003;30:225-244.
10. Lindstrom M, Hanson BS, Ostergren P-O. Socioeconomic differences in leisure-time
physical activity: the role of social participation and social capital in shaping health
related behaviour. Soc Sci Med. 2001;52:441-451.
11. Estabrooks PA, Lee RE, Gyurcsik NC. Resources for physical activity participation:
does availability and accessibility differ by neighbourhood socioeconomic status? Ann
Behav Med. 2003;25:100-104.
12. Papadopoulou SK, Papadopoulou SD, Zerva A et al. Health status and socioeconomic
factors as determinants of physical activity level in the elderly. Med Sci Monit. 2003;9:
CR79-83.
13. Lawlor DA, Taylor M, Bedford C, Ebrahim S. Is housework good for health? Levels of
physical activity and factors associated with activity in elderly women. Results from
the British Women’s Heart and Health Study. J Epidemiol Community Health. 2002;56:
473-478.
14. Leslie E, Cerin E, Gore CJ et al. Gender, age and educational-attainment differences
in Australian adults’ participation in vigorous sporting and fitness activities. J Phys
Activity Health. 2004;1:377-388.
15. Kaplan MS, Huguet N, Newsom JT, MacFarland BH. Characteristics of physically
inactive older adults with arthritis: results of a population-based study. Prev Med.
2003;37:61-67.
16. Pereira MA, Kriska AM, Collins VR et al. Occupational status and cardiovascular
disease risk factors in the rapidly developing, high-risk population of Mauritius. Am J
Epidemiol. 1998;148:148-159.
17. Humpel N, Owen N, Iverson D et al. Perceived environment attributes, residential
location, and walking for particular purposes. Am J Prev Med. 2004;26:119-125.
18. Ross CE. Walking, exercising, and smoking: does neighbourhood matter? Soc Sci Med.
2000;51:265-274.
19. Salmon J, Bauman A, Crawford D et al. The association between television viewing
and overweight among Australian adults participating in varying levels of leisure-time
physical activity. Int J Obes. 2000;24:600-606.
20. Australian Bureau of Statistics. National Health Survey: Users Guide (Cat No.
4363.0.55.001). Canberra: Australian Government Publishing Service, 2003.
21. World Health Organisation. Global strategy on diet, physical activity and health. Obesity
and overweight. Vol. 2004: World Health Organization, 2003.
22. Australian Bureau of Statistics. Socio-economic indexes for areas (2912.0). Canberra:
Australian Government Publishing Service, 1991
23. StataCorp. Stata Base Reference Manual, Volume 1. College Station, TX: Stata Cor-
poration, 2003
24. StataCorp. Stata Statistical Software: Release 8.2. College Station, TX: Stata Corpora-
tion, 2004
25. Humpel N, Marshall AL, Leslie E et al. Changes in neighbourhood walking are related
to changes in perceptions of environmental attributes. Annals of Behavioral Medicine.
2004;27:60-67
26. Kawachi I, Kennedy BP, Wilkinson RG. Crime: social disorganization and relative
deprivation. Soc Sci Med. 1999;48:719-731.
 Socioeconomic Predictors of a Sedentary Lifestyle 101

27. Bolin K, Lindgren B, Lindstrom M, Nystedt P. Investments in social capital—implica-

tions of social interactions for the production of health. Soc Sci Med. 2003;56:2379-
2390.
28. Shephard RJ. Benefits of sport and physical activity for the disabled: implications for
the individual and for society. Scand J Rehabil Med. 1991;23:51-59.
29. Brown WJ, Trost SG. Life transitions and changing physical activity patterns in young
women. Am J Prev Med. 2003;25:140-143.
30. Wilcox S, Castro C, King AC et al. Determinants of leisure time physical activity in
rural compared with urban older and ethnically diverse women in the United States. J
Epidemiol Community Health. 2000;54:667-672.
31. Dixon T, Walters A-M. A growing problem: trends and patterns in overweight and
obesity among adults in Australia, 1980 to 2001. Canberra: Australian Institute of
Health and Welfare, 2003.
32. DiPietro L. Physical activity, body weight, and adiposity: an epidemiological perspec-
tive. Exerc Sport Sci Rev. 1995;23:275-303.
33. I-Min L, Paffenbarger R. Associations of light, moderate, and vigorous intensity physical
activity with longevity: The Harvard Alumni Health Study. Am J Epidemiol. 2000;151:
293-299.