Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

The socioeconomic gradient and chronic illness and associated risk factors in Australia

The socioeconomic gradient and chronic illness and associated risk factors in Australia Objective: To examine the prevalence of major chronic diseases and their risk factors in different socioeconomic groups in the Australian population, in order to highlight the need for public policy initiatives to reduce socioeconomic inequalities in health. Methods: Data were provided by the Australian Bureau of Statistics (ABS) from the 2001 National Health Survey (NHS) for selected chronic diseases and associated risk factors. Conditions selected were those, which form the National Health Priority Area (NHPA) conditions (other than injury, which has not been included in this paper, with its focus on chronic disease); plus other 'serious' chronic conditions, in line with the classification developed by Mathers; and for which sufficient cases were available for analysis by socioeconomic status. Indirectly age-standardised prevalence rates were calculated by broad age group for Australia and for five groups of socioeconomic status; rate ratios were calculated to show variations in prevalence between these groups. Results: Significant socioeconomic inequalities were evident for many of the major chronic diseases; the largest was for diabetes mellitus (at ages 25 to 64 years); and for many diseases, there was also a strong, continuous socioeconomic gradient in the rates. Circulatory system diseases (in particular, hypertensive disease) and digestive system diseases also exhibited a strong differential in the 25 to 64 year age group. In the 65 years and over age group, the strongest inequalities were evident for mental and behavioural problems, diabetes (with a continuous socioeconomic gradient in rates) and respiratory system diseases. A number of risk factors for chronic diseases, namely self-reported smoking, alcohol misuse, physical inactivity and excess weight showed a striking association with socioeconomic status, in particular for people who were smokers and those who did not exercise. Conclusion: This analysis shows that the prevalence of chronic disease varies across the socioeconomic gradient for a number of specific diseases, as well as for important disease risk factors. Therefore, any policy interventions to address the impact of chronic disease, at a population level, need to take into account these socioeconomic inequalities. Page 1 of 8 (page number not for citation purposes) Australia and New Zealand Health Policy 2004, 1:8 http://www.anzhealthpolicy.com/content/1/1/8 grouping residential locations according to socioeco- Background As in other developed countries, chronic diseases in Aus- nomic criteria. A number of such studies have docu- tralia are major contributors to the extent of illness, disa- mented substantial variations in mortality for different bility and premature mortality in the population. They are age groups [11-17]. estimated to make up the greatest proportion of the bur- den of disease, mental problems and injury for the popu- However, to date, there have been fewer studies that have lation as a whole (about 80%), and for particular sub- examined socioeconomic inequalities in chronic disease population groups [1]. prevalence in those still living (analyses of hospital admis- sions for chronic conditions have been published, but not Chronic diseases are exemplified by having multifactorial of prevalence) [15]. One of the earliest was undertaken by aetiologies, including common disease risk factors and Broadhead, who analysed data on morbidity and social determinants; significant latency periods and protracted status from the 1977–78 Australian Health Survey (ABS), clinical courses; and are seldom cured completely [2,3]. and found that men in lower status occupations tended to Causal factors interact together at an individual and at a suffer a higher age-standardised rate of self-reported population level to determine the degree of disease bur- chronic conditions and days of reduced activity; the pic- den and illness, and unhealthy risks can be passed on ture for women appeared less clear [18]. Lee et al found through families, communities, and populations follow- that low income males were more likely to report mental ing demographic gradients [4]. At different life stages, health problems, chronic symptoms and acute symptoms common risk factors and determinants include poor intra- than their high income counterparts [19]. Similar findings uterine conditions; stress, violence and traumatic experi- were reported in other studies, and risk factors associated ences; educational disadvantage; inadequate living envi- with chronic diseases also were also associated with low ronments that fail to promote healthy lifestyles; poor diet income [15,19-21]. and lack of exercise; alcohol misuse and tobacco smoking [5,30]. Risk factors are also increasingly more prevalent in The work of Mathers is significant for its systematic docu- areas of low socioeconomic status and in communities mentation of health inequalities among working aged characterised by low levels of educational attainment; Australians (25 to 64 years) in the late 1980s. He exam- high levels of unemployment; substantial levels of dis- ined mortality, disability, disease groups, specific dis- crimination, interpersonal violence and exclusion; and eases, self-perceived health, risk factors, health service use poverty. There is a higher prevalence of such factors and use of preventive screening services, using data from among Indigenous communities, and other socioeco- the 1989–90 National Health Survey (NHS) [15]. Mathers nomically disadvantaged Australians [5,6]. found that there were no clear gradients of chronic, recent or minor illness with level of socioeconomic disadvantage The inequalities in health observed across populations are of area, although there were some specific health status many – some of them are inevitable and others, unneces- indicators (self-reported health, reduced activity, unhap- sary and unfair. Those inequalities that are potentially piness) and certain risk factors (inactivity, smoking and avoidable are deemed 'inequitable' [7]. Despite signifi- alcohol use) that were reported more frequently by those cant medical advances and improved public health in in the more disadvantaged quintiles [15]. recent decades, socioeconomically disadvantaged com- munities continue to suffer an unequal burden of illness, Current information on inequalities in health other than premature death and disability. Therefore, the study of mortality is limited in Australia because of a dearth of socioeconomic inequalities in chronic diseases and condi- suitable data collections. However, the release of data tions and in risk factors is important and necessary. This is from the 2001 NHS provides an opportunity to examine particularly so, if we wish to develop more effective policy the prevalence of self-reported chronic disease in Australia mechanisms for preventing and intervening earlier in the and the way in which this impacts on different socioeco- progression of chronic diseases and their associated risk nomic groupings within the population. factors across the diverse Australian population, and to reduce some of the existing health inequities. Results Information for a selection of chronic diseases is shown in Our approach Table 1. Diseases were included on the basis of either high There have been a number of studies published in Aus- prevalence or their contribution to the burden of disease. tralia on socioeconomic inequalities in mortality from various chronic diseases and conditions. The earlier ones The main findings are: were analysed using information on occupation recorded on the death certificate [8-10]. An alternative approach The largest differential between those in the most well off has been to examine variations in mortality rates by and those in the most disadvantaged areas was for Page 2 of 8 (page number not for citation purposes) Australia and New Zealand Health Policy 2004, 1:8 http://www.anzhealthpolicy.com/content/1/1/8 Table 1: Inequality in prevalence of selected chronic diseases , 2001 2 3 Age group (years) and chronic disease Rate Rate ratio by quintile of socioeconomic disadvantage of area First Second Third Fourth Fifth 0–14 Mental and behavioural problems 6 596 1.00 1.04 1.10 1.12 1.52** Respiratory system 21 807 1.00 1.07 1.05 1.11 0.99 Asthma 13 363 1.00 1.10 1.12 1.25* 1.12 15–24 Mental and behavioural problems 10 284 1.00 1.02 0.97 1.08 1.28 Respiratory system 33 373 1.00 1.04 1.12 1.09 1.00 Asthma 16 263 1.00 0.82 1.14 1.02 1.00 6 6 6 6 6 Bronchitis/emphysema 1 701 1.00 1.32 1.66 1.94 1.97 Musculoskeletal system 19 088 1.00 1.11 1.00 1.08 0.94 25–64 Diabetes mellitus 2 234 1.00 1.37 1.67* 1.72* 2.28*** Mental and behavioural problems 11 093 1.00 1.05 1.20* 1.36*** 1.67*** Circulatory system 17 491 1.00 1.04 0.97 1.15* 1.28*** Hypertensive disease 9 751 1.00 1.12 1.01 1.24* 1.54*** Respiratory system 32 964 1.00 1.00 0.99 0.99 1.01 Asthma 10 393 1.00 1.10 0.99 1.19* 1.14 Bronchitis/emphysema 3 429 1.00 0.97 1.14 1.55** 1.70*** Digestive system 8 074 1.00 1.03 1.07 1.12 1.37*** Musculoskeletal system 39 840 1.00 1.10* 1.16*** 1.16*** 1.22*** 65 & over Diabetes mellitus 8 981 1.00 1.13 1.14 1.52* 1.56* Mental and behavioural problems 7 222 1.00 1.21 1.62* 1.67* 1.56* Circulatory system 56 592 1.00 1.09 1.06 1.10 1.19* Respiratory system 31 442 1.00 1.03 0.87 0.95 1.22* Musculoskeletal system 63 669 1.00 1.02 1.06 1.03 1.08 Survey respondents can report more than one disease. Rate is the number of persons per 100,000 population reporting the disease. The extent of any inequality is shown by the rate ratio, which expresses the ratio of the rate in each quintile to the rate in Quintile 1 (the most advantaged areas, with a rate ratio of 1.00); rate ratios differing significantly from 1.0 are shown with * p < 0.05; ** p < 0.01; *** p < 0.001. Information about these age groups were collected by proxy, using parental report. Information may be based on self-diagnosis, rather than diagnosis by a health practitioner. Indicates rate ratio based on estimates with a Relative Standard Error of between 25% and 50% and should be used with caution. Includes diseases of the connective tissue. Source: National Health Survey, ABS 2002 diabetes mellitus at ages 25 to 64 years, with the preva- Circulatory system diseases (in particular, hypertensive lence in the most disadvantaged areas being just over two disease) and digestive system diseases also exhibit a strong and a quarter times (a rate ratio of 2.28) the prevalence for differential in the 25 to 64 year age group (statistically sig- the least disadvantaged; there is also a strong, continuous nificant differentials of 28% and 54%, respectively). gradient in the rates, with the rate ratios in each of the third to fifth quintiles statistically significant. In the 65 years and over age group, the strongest differen- tials were evident for mental and behavioural problems (a There was a statistically significant differential of 67% at statistically significant 56%), diabetes (with a continuous ages 25 to 64 years, with a strong, continuous gradient, in gradient in rates, statistically significant in quintile3 four the prevalence of self-reported mental and behavioural and five) and respiratory system disease (a statistically sig- problems across the socioeconomic gradient; differentials nificant 22%). (also statistically significant) in the 0 to 14 year and 65 years and over age groups were 52% and 56%, Asthma accounted for almost two thirds of the rate of respectively. reporting of respiratory system disease in the 0 to 14 year Page 3 of 8 (page number not for citation purposes) Australia and New Zealand Health Policy 2004, 1:8 http://www.anzhealthpolicy.com/content/1/1/8 Table 2: Inequality in prevalence of selected health risk factors, 18–64 years, 2001 2 3 Health risk factors Rate Rate ratio by quintile of socioeconomic disadvantage of area First Second Third Fourth Fifth Current smokers - Male 30 582 1.00 1.40*** 1.55*** 1.71*** 1.95*** - Female 24 009 1.00 1.29*** 1.34*** 1.48*** 2.00*** - Persons 27 275 1.00 1.35*** 1.45*** 1.61*** 1.96*** Alcohol – High risk - Males 6 976 1.00 1.09 1.26* 1.26* 1.45*** - Females 2 127 1.00 0.59* 0.94 0.76 0.87 - Persons 4 537 1.00 0.93 1.16 1.12 1.22* Did not exercise - Males 28 772 1.00 1.20** 1.36*** 1.52*** 1.68*** - Females 28 220 1.00 1.19** 1.29*** 1.35*** 1.65*** - Persons 28 494 1.00 1.20*** 1.32*** 1.43*** 1.66*** Underweight females 12 675 1.00 0.89 0.83* 0.72*** 0.91 Overweight/obese - Males 54 701 1.00 1.09* 1.11* 1.04*** 1.00 - Females 37 004 1.00 1.09 1.21*** 1.16** 1.17** - Persons 45 798 1.00 1.09** 1.15*** 1.09** 1.06 Survey respondents can be shown under more than one type of risk factor. Rate is the number of persons per 100,000 population estimated to be at risk from the health risk factor. The extent of any inequality is shown by the rate ratio, which expresses the ratio of the rate in each quintile to the rate in Quintile 1 (the most advantaged areas, with a rate ratio of 1.00); rate ratios differing significantly from 1.0 are shown with * p < 0.05; ** p < 0.01; *** p < 0.001. Source: National Health Survey, ABS 2002 age group, almost half in the 15 to 24 year age group, and level measures of SES were used [28]. Thus, chronic dis- for about a third of the rate in the 25 to 64 year age group. ease inequalities in the wider population by SES are likely to be larger than those reported in this study. In addition, The NHS also included data on a number of risk factors the exclusion of the 'sparsely settled' areas of Australia in for chronic diseases, namely self-reported smoking, alco- NHS data collection results in the omission of data from hol misuse, physical inactivity and excess weight. A a high percentage of Indigenous people, who are the pop- number of these risk factors show a striking association ulation group with the poorest health. with socioeconomic status, in particular for people who are smokers and those who did not exercise, with contin- Discussion uous gradients and highly elevated rates of statistical sig- Our analysis indicates that socioeconomic inequalities in nificance (Table 2). The differences in male and female the prevalence of chronic diseases and their concomitant rates are also of interest. It was only for underweight risk factors are evident across the Australian population. females, and for the risk factor of high-risk alcohol con- However, the diseases with substantial disparities across sumption by females, that the socioeconomic gradient the socioeconomic quintiles are different, for different was reversed. stages in the life course. Although these results cannot be directly compared with those of previous studies, because It is important to note that the inequalities reported above of definitional and methodological differences, the recur- relate to the health of those people living in a geographic ring finding of inequalities for chronic disease morbidity area and to the overall level of socioeconomic disadvan- and risk factor prevalence across the socioeconomic gradi- tage of that area. Most areas will contain varying levels of ent remains a significant concern. individual socioeconomic disadvantage and, to the extent that the poorer health is associated with individual The burden in the Australian population attributable to economic circumstances and living conditions rather than socioeconomic inequality is large, and has far-reaching communal environment, the inequalities will understate implications in terms of unnecessary disability and suffer- the true differences in health status according to socioeco- ing, the loss of potentially economically productive mem- nomic disadvantage [15]. bers of society, and increased costs for the health and social care systems [35]. Despite the expenditure of mil- Furthermore, there are limitations to the use of area-based lions of dollars to prevent and reduce the prevalence of measures of SES. Due to misclassification error (i.e. ascrib- chronic diseases and their risk factors, these inequities ing area-SES to individuals), estimates of difference across have persisted. However, the situation in Australia is by the quintiles will be smaller than if data on individual- no means unique, for inequalities in these diseases and Page 4 of 8 (page number not for citation purposes) Australia and New Zealand Health Policy 2004, 1:8 http://www.anzhealthpolicy.com/content/1/1/8 their risk factors have been observed for most of the devel- Policy-makers who wish to address socioeconomic ine- oped countries in which they have been studied [26]. qualities in health may favour one of the following approaches. Some view the impact of socioeconomic dis- What should we be doing differently? There is a growing advantage on those groups with the poorest health in the body of knowledge that will help to provide direction for population, such as Aboriginal people and Torres Strait developing policies to reduce inequities across the popu- Islanders, as the priority policy goal. Others identify the lation. The socioeconomic environment is a powerful and gap between the health of those groups at the outer ends potentially modifiable factor, and public policy is a key of the socioeconomic hierarchy (those with the poorest instrument to improve this environment, particularly in health and those with best health), and see the narrowing areas such as housing, taxation and social security, work of the gap as the goal. Others prefer to focus on the socio- environments, urban design, pollution control, educa- economic gradient in health that runs across the whole tional achievement, and early childhood development population [31]. [34]. Graham has identified that the last approach widens the However, attention must be paid to the nature of any policy debate in three ways [31]. Firstly, it looks for the action that is taken, to ensure that social and economic causes of health inequality in the systematic differences in inequalities are not increased. Some programs, by their life chances and opportunities, living standards and life- very success, can increase inequality by widening the gap styles that are associated with people's unequal positions between groups in the population; for example, such pro- across the socioeconomic hierarchy, and for the pathways grams may be more attractive to those who are already through which they influence health [31,36]. Secondly, as healthier, or not as effective for certain groups with poorer a result, addressing health inequalities becomes a popula- health, less education or more stressful lives. In one smok- tion-wide goal that includes every citizen [31]. Thirdly, ing cessation initiative, it was found that the prevalence of 'reducing health gradients' provides a more comprehen- smoking decreased predominately in those adults with sive policy approach: one that encompasses 'remedying higher education, thus increasing the existing difference disadvantages' and 'narrowing health gaps' within the with those who were more disadvantaged [37]. While broader goal of 'equalising health chances across all the smoking prevalence in Australia has reduced considerably socioeconomic groups' [31]. over the last 20 years, attributes such as lower education and occupational status, unemployment, rented housing, She also observes that, "improving the health of poor and living in disadvantaged areas reflect a higher proba- groups and improving their position relative to other bility of reporting tobacco expenditure [32]. As a result, groups are necessary elements in a strategy to reduce the the tax revenue from the sale of tobacco products is being socioeconomic gradient. However, neither is sufficient on disproportionately drawn from the poorest households its own. To reduce the socioeconomic gradient, health in and represents a greater proportion of their household other socioeconomic groups also needs to improve at a budget [32]. faster rate than in the highest socioeconomic group. Thus, policies to ameliorate health disadvantage, to close health It is also evident that the ways in which systems such as gaps and to reduce health gradients need to be pursued education and health are delivered and structured can together, and not at the expense of each other" [31]. increase existing inequality. For example, schooling can be a way of addressing inequality and also a way of repro- There is also an urgent need to make health inequalities a ducing it. It has been suggested that there are two goals for research priority for each stage of the life course – not just a social justice program in education: to work to eliminate to monitor the size and extent of the disparities but also the contribution that the education system makes to the to undertake research that will find preventive approaches production over time of social inequality in general; and and further policy interventions that will be effective in to maximise the positive contributions that the education reducing them, and that are likely to be implemented by system makes to reducing social inequality [33]. There- governments and communities. fore, different approaches and mixes of policies and pro- grams must be mounted to address inequalities. These Conclusions approaches may include more precise targeting, but also Clearly, any moves to address the impact of chronic dis- greater attention to community-based dimensions of ease at a population level must take into account socioe- 'interdependence' between individual behaviours, key conomic inequalities in prevalence. More research is determinants, and community and institutional needed to determine which approaches are effective and resources. why others have failed to have the desired impact, partic- ularly for those who are from socioeconomically disad- vantaged areas. Finally, although rates are generally Page 5 of 8 (page number not for citation purposes) Australia and New Zealand Health Policy 2004, 1:8 http://www.anzhealthpolicy.com/content/1/1/8 highest at the oldest ages, the development of risk factors the classification developed by Mathers [15]; and for for many chronic diseases occurs early in life, and thus, it which sufficient cases were available for analysis by five is essential those health inequities are addressed right groups of socioeconomic disadvantage of area (see below across the life course. for details of the way these groups were constructed). Methods The risk factors used by the ABS were those identified for Data sources the NHPA conditions [27]. The ABS conducts the National Health Survey (NHS) on a regular basis, most recently in 2001 [22]. The NHS collects The ABS has coded conditions reported by respondents to information from approximately 26,900 people from all output disease categories based on ICD-10. Conditions States and Territories living in private dwellings, selected described as 'chronic' in this article include those long- at random using a multi-stage area sample of private term conditions reported in the NHS, which are com- dwellings. The survey is undertaken across much of Aus- monly recognised by health practitioners as chronic dis- tralia, but excludes the 'sparsely settled' areas, which com- eases [23]. The risk factor for 'high risk due to alcohol' prised less than 1% of the non-Indigenous population reflects the National Health and Medical Research and 25% of the Indigenous population at the 2001 Cen- Council's risk levels for harm in the long term from alco- sus: a separate Australia-wide survey of the health of hol consumption [24]. The risk factors for overweight and Indigenous people, also conducted in 2001, surveyed underweight were calculated from self-reported height these sparsely settled areas. and weight information and grouped to reflect World Health Organization (WHO) guidelines [25]. The survey includes self-reported details of health condi- tions (both acute and long term) and major risk factors, as Given the policy importance of the NHPAs, the 2001 NHS well as demographic and socioeconomic information questionnaire underwent significant revisions to more about the survey respondent. Respondents were asked if precisely capture information on several of the NHPAs. they had been told by a doctor or nurse that they had Consequently, while the quality of the information on asthma, cancer, heart and circulatory conditions, and/or NHPAs has been improved from the 1995 NHS, the diabetes. These conditions, together with injuries and degree of comparability with previous surveys has been mental health, form the NHPAs [27]. However, for long somewhat compromised for many of the major health term mental health problems, respondents were not asked conditions. Some specific conditions (e.g., diabetes) whether they had been told by a doctor or nurse that they appear to be comparable between the 1989–90 and 2001 had any mental health problems; thus, the responses may surveys, however, for most groups of conditions based on be based on self-diagnosis, rather than diagnosis by a ICD chapter headings (e.g., all circulatory) the ABS advise health practitioner [22]. Respondents were also asked a that the combined effect of major conceptual changes as series of questions about other specific, non-NHPA, con- well as major classification changes between the 1989–90 ditions, covering eye and sight problems, ear and hearing and 2001 surveys would make direct comparisons very problems, and arthritis, rheumatism and gout. They were difficult. This analysis is therefore restricted to the 2001 then shown a series of three prompt cards (two with con- data. ditions listed, while the third contained more general Measurement of socioeconomic status descriptions of condition types) and asked whether they had any of the conditions shown or conditions similar to The socioeconomic status (SES) of the address of resi- those shown or described. In each of these cases, details dence of each survey respondent is available at the Census were recorded for conditions reported as current at the Collection District (CD) level and was added to the NHS time of the survey; respondents were also asked whether file, as was the quintile of socioeconomic disadvantage of the condition had lasted, or was expected to last, for six area into which that CD fell at the Census. The measure months or more. Information was gathered directly from used to allocate CDs to quintiles was the 1996 Census individuals aged 15 years and older. For children up to the Index of Relative Socio-Economic Disadvantage (IRSD). age of 15 years, information was provided by proxy, from The IRSD is one of five Socio-Economic Indexes for Areas a parent or guardian. produced by the ABS using Principal Components Analy- sis. It summarises information available from variables The particular conditions for which data were requested collected in the 1996 Population Census including those from the ABS for this analysis were: related to education, occupation, and income. The varia- bles are expressed as percentages of the relevant the NHPA conditions (other than injury, which has not population. been included in this paper, with its focus on chronic dis- ease); plus other 'serious' chronic conditions, in line with Page 6 of 8 (page number not for citation purposes) Australia and New Zealand Health Policy 2004, 1:8 http://www.anzhealthpolicy.com/content/1/1/8 The NHS records were aggregated to the quintiles derived ratio of the rate in each quintile to the rate in Quintile 1 from the Census data, where Quintile 1 comprises the (the most advantaged areas, with a rate ratio of 1.00). CDs with the highest IRSD scores (highest socioeconomic status, or most advantaged, areas) and Quintile 5 com- Competing interests prises the CDs with the lowest IRSD scores (lowest socio- The author(s) declare that they have no competing economic status, or most disadvantaged areas). Each interests. quintile comprises approximately 20% of CDs. Authors' contributions The ABS produced the estimates of the number of people JG conceived the study, and was responsible for its design with chronic diseases and risk factors by quintile. The and coordination. ST participated in the design of the method used resulted in the production of quintiles of study and performed the statistical analysis. DH partici- varying sizes, ranging from 17.4% of the population in pated in the drafting of the manuscript. All authors read Quintile 5 (most disadvantaged areas) to 22.8% in Quin- and approved the final manuscript. tile 2 and 21.1% in Quintile 1 (most advantaged areas). This is a differential of over five percentage points Acknowledgements The Australian Government Department of Health and Ageing funded this between Quintile 5 and Quintile 2, or 1,027,030 fewer research. The data were provided by the ABS. people in the most disadvantaged areas when compared with Quintile 2 (and 708,980 fewer in Quintile 5 than in References Quintile 1). The effect of this lack of precision on the 1. Mathers CD, Vos ET, Stevenson CE, Begg SJ: The Australian Bur- results by quintile is not known. Although, in part, the dif- den of Disease Study: measuring the loss of health from dis- ference arises as a result of the method used (that is, that eases, injuries and risk factors. Med J Aust 2000, 172:592-596. 2. Australian Institute of Health and Welfare: Chronic diseases and associ- the quintiles are based on the Census population, and ated risk factors in Australia, 2001 Canberra; 2002. [AIHW Catalogue applied to the NHS population which has a different pro- No. PHE 33] 3. Thacker S, Stroup D, Rothenberg R: Public health surveillance for file), it may also reflect different response rates to the sur- chronic conditions: a scientific basis for decisions. Statistics in vey from different socioeconomic groups. Medicine 1995, 14:629-641. 4. Ackland M, Choi BC, Puska P: Rethinking the terms non-com- municable disease and chronic disease. J Epidemiol Community The NHS includes other measures that could potentially Health 2003, 57:838-9. be used to measure socioeconomic status. For example, 5. National Public Health Partnership (NPHP): Preventing Chronic Disease: A Strategic Framework Melbourne, Victoria; 2001. income reporting in ABS surveys is known to be incom- 6. Mooney G: Inequity in Australian health care: how do we plete, in particular for low income groups and, in the progress from here? Aust NZ J Pub Health 2003, 27:267-270. 2001 NHS, income was only collected for the respondent 7. Whitehead M: The concepts and principles of equity in health. Int J Health Serv 1992, 22:429-445. and their spouse. Information on education is also avail- 8. McMichael AJ, Hartshorne JM: Mortality risks in Australian men able, but a single measure has yet to be agreed. Problems by occupational groups, 1968–1978: variations associated in using education include that it is typically completed with differences in drinking and smoking habits. Med J Aust 1982, 1:253-6. early in adulthood; it captures neither differential on-the- 9. Taylor R, Herman H, Preston G: Occupation and mortality in Australian job training and other career investments made by indi- working age males, 1975–77 Health Commission of Victoria and Department of Social and Preventive Medicine, Monash University: viduals with similar levels of formal schooling, nor the Melbourne; 1983. volatility in economic status during adulthood that has 10. McMichael AJ: Social class (as estimated by occupational pres- been shown to have adverse implications for health [29]. tige) and mortality in Australian males in the 1970s. Commu- nity Health Stud 1985, 9:20-30. The age structure of the population may also influence the 11. Siskind V, Najman J, Copeman R: Socioeconomic status and mor- indicator: an older population generally has lower educa- tality: a Brisbane area analysis. Community Health Stud 1987, 11:15-23. tion levels than a younger population due to improved 12. South Australian Health Commission (SAHC): A Social Health Atlas of access to education over time. South Australia. Adelaide 1990. 13. McNeil J, King R, Jennings G, Powles J, eds: A textbook of preventative medicine Melbourne: Edward Arnold; 1990. Analysis 14. National Health Strategy: Enough to make you sick: how income Chronic disease and risk factor rates are expressed as rates and environment affect health. [National Health Strategy Research per 100,000 population, indirectly age-standardised. The Paper No.1]. Melbourne 1992. 15. Mathers C: Health inequalities among adult Australians aged rates were also calculated by age although, given the small 25–64 years. In [Australian Institute of Health and Welfare Health Mon- sample size of the NHS, only broad age groups (0–14 itoring Series No. 1] Canberra: AGPS; 1994. 16. Glover J, Harris K, Tennant S: A Social Health Atlas of Australia 2nd edi- years, 15–24 years, 25–64 years and 65 years and over) tion. Adelaide: Openbook; 1999. were available. The standard population and quintile 17. Turrell G, Mathers C: Socioeconomic inequalities in all-cause populations are the weighted 2001 survey populations and specific-cause mortality in Australia: 1985–1987 and 1995–1997. Int J Epidemiol 2001, 30:231-239. from the NHS. The extent of any differential between the 18. Broadhead P: Social status and morbidity in Australia. Commu- quintiles is shown by the rate ratio, which expresses the nity Health Stud 1985, 9:87-98. 19. Lee S-H, Smith L, d'Espaignet E, Thompson N: Health Differentials for Working Age Australians Canberra: Australian Institute of Health; 1987. Page 7 of 8 (page number not for citation purposes) Australia and New Zealand Health Policy 2004, 1:8 http://www.anzhealthpolicy.com/content/1/1/8 20. Broom D: The social distribution of illness: Is Australia more equal? Soc Sci Med 1984, 18:909-917. 21. Dobson A, Gibberd R, Leeder S, O'Donnell D: Occupational dif- ferences in ischaemic heart disease mortality and risk fac- tors in Australia. Am J Epidemiol 1985, 122:282-90. 22. Australian Bureau of Statistics (ABS): National Health Survey: Summary of Results Canberra: ABS; 2002. [ABS Cat. No. 4364.0] 23. Rothenberg R, Koplan J: Chronic disease in the 1990s. Annu Rev Public Health 1990, 11:267-96. 24. National Health and Medical Research Council (NHMRC): Australian Alcohol Guidelines: Health Risks and Benefits Canberra: AusInfo; 2001. 25. World Health Organization (WHO): Obesity: Preventing and Managing the Global Epidemic – Report of a WHO Consultation. Geneva 2000. [WHO Technical Report Series, No. 894] 26. Beaglehole R, Yach D: Globalisation and the prevention and control of non-communicable disease: the neglected chronic diseases of adults. Lancet 2003, 362:903-08. 27. Australian Government Department of Health and Ageing: National Health Priority Areas. [http://www.health.gov.au/pq/nhpa/]. accessed 27 July 2004 28. Hyndman JC, Holman CSD, Hockey RL, Donovan RJ, Corti B, Rivera J: Misclassification of social disadvantage based on geograph- ical areas: comparison of postcode and collector's district analyses. International Journal of Epidemiology 1995, 24:165-176. 29. McDonough P, Duncan DJ, Williams DR, House J: Income dynam- ics and adult mortality in the US, 1972–1989. American Journal of Public Health 1997, 18:341-378. 30. Williams DR: Race, socioeconomic status, and health: the added effects of racism and discrimination. In In: Socioeconomic status and health in industrial nations: social, psychological, and biological pathways Edited by: Adler NE, Marmot M, McEwen S, Stewart J. New York: Annals of the New York Academy of Science; 1999:173-188. 31. Graham H: Tackling Inequalities in Health in England: Reme- dying Health Disadvantages, Narrowing Health Gaps or Reducing Health Gradients? Journal of Social Policy 2004, 33(1):115-131. 32. Siahpush M, Borland R, Scollo M: Smoking, socioeconomic status, and household expenditure on food, alcohol, gambling and insurance The Can- cer Council of Victoria: Carlton South; 2003. 33. Connell RW, White V, Johnston K, eds: Running Twice As Hard: the dis- advantaged schools program in Australia Geelong: Deakin University; 34. Hetzel D, Page A, Glover J, Tennant S: Inequality in South Australia: The Evidence Volume 1. Adelaide: Department of Health, SA; 2004. 35. Centre for South Australian Economic Studies (CSAES): The Impact of Socio-Economic and Locational Disadvantage on Health Outcomes and Cost. In Social Justice Research Program into Locational Disadvantage – Report no. 10 Department of Health, Housing, Local Government and Community Services: AGPS, Canberra; 1993. 36. Davey Smith G, Gunnell D, Ben-Shlomo Y: Life-course approaches to socioeconomic differentials in cause-specific adult mortal- ity'. In in Poverty, Inequality and Health Edited by: Leon D, Walt G. Oxford: Oxford University Press; 2001. 37. Osler M, Gerdes LU, Davidsen M, Brønnum-Hansen H, Madsen M, Jørgensen T, Schroll M: Socioeconomic status and trends in risk factors for cardiovascular diseases in the Danish MONICA population, 1982–1992. Journal of Epidemiology and Community Health 2000, 54:108-13. Publish with Bio Med Central and every scientist can read your work free of charge "BioMed Central will be the most significant development for disseminating the results of biomedical researc h in our lifetime." Sir Paul Nurse, Cancer Research UK Your research papers will be: available free of charge to the entire biomedical community peer reviewed and published immediately upon acceptance cited in PubMed and archived on PubMed Central yours — you keep the copyright BioMedcentral Submit your manuscript here: http://www.biomedcentral.com/info/publishing_adv.asp Page 8 of 8 (page number not for citation purposes) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Australia and New Zealand Health Policy Springer Journals

The socioeconomic gradient and chronic illness and associated risk factors in Australia

Download PDF
8 pages

Loading next page...
 
/lp/springer-journals/the-socioeconomic-gradient-and-chronic-illness-and-associated-risk-Kv6GASbmrR

References (36)

Publisher
Springer Journals
Copyright
Copyright © 2004 by Glover et al; licensee BioMed Central Ltd.
Subject
Medicine & Public Health; Public Health; Social Policy
ISSN
1743-8462
eISSN
1743-8462
DOI
10.1186/1743-8462-1-8
Publisher site
See Article on Publisher Site

Abstract

Objective: To examine the prevalence of major chronic diseases and their risk factors in different socioeconomic groups in the Australian population, in order to highlight the need for public policy initiatives to reduce socioeconomic inequalities in health. Methods: Data were provided by the Australian Bureau of Statistics (ABS) from the 2001 National Health Survey (NHS) for selected chronic diseases and associated risk factors. Conditions selected were those, which form the National Health Priority Area (NHPA) conditions (other than injury, which has not been included in this paper, with its focus on chronic disease); plus other 'serious' chronic conditions, in line with the classification developed by Mathers; and for which sufficient cases were available for analysis by socioeconomic status. Indirectly age-standardised prevalence rates were calculated by broad age group for Australia and for five groups of socioeconomic status; rate ratios were calculated to show variations in prevalence between these groups. Results: Significant socioeconomic inequalities were evident for many of the major chronic diseases; the largest was for diabetes mellitus (at ages 25 to 64 years); and for many diseases, there was also a strong, continuous socioeconomic gradient in the rates. Circulatory system diseases (in particular, hypertensive disease) and digestive system diseases also exhibited a strong differential in the 25 to 64 year age group. In the 65 years and over age group, the strongest inequalities were evident for mental and behavioural problems, diabetes (with a continuous socioeconomic gradient in rates) and respiratory system diseases. A number of risk factors for chronic diseases, namely self-reported smoking, alcohol misuse, physical inactivity and excess weight showed a striking association with socioeconomic status, in particular for people who were smokers and those who did not exercise. Conclusion: This analysis shows that the prevalence of chronic disease varies across the socioeconomic gradient for a number of specific diseases, as well as for important disease risk factors. Therefore, any policy interventions to address the impact of chronic disease, at a population level, need to take into account these socioeconomic inequalities. Page 1 of 8 (page number not for citation purposes) Australia and New Zealand Health Policy 2004, 1:8 http://www.anzhealthpolicy.com/content/1/1/8 grouping residential locations according to socioeco- Background As in other developed countries, chronic diseases in Aus- nomic criteria. A number of such studies have docu- tralia are major contributors to the extent of illness, disa- mented substantial variations in mortality for different bility and premature mortality in the population. They are age groups [11-17]. estimated to make up the greatest proportion of the bur- den of disease, mental problems and injury for the popu- However, to date, there have been fewer studies that have lation as a whole (about 80%), and for particular sub- examined socioeconomic inequalities in chronic disease population groups [1]. prevalence in those still living (analyses of hospital admis- sions for chronic conditions have been published, but not Chronic diseases are exemplified by having multifactorial of prevalence) [15]. One of the earliest was undertaken by aetiologies, including common disease risk factors and Broadhead, who analysed data on morbidity and social determinants; significant latency periods and protracted status from the 1977–78 Australian Health Survey (ABS), clinical courses; and are seldom cured completely [2,3]. and found that men in lower status occupations tended to Causal factors interact together at an individual and at a suffer a higher age-standardised rate of self-reported population level to determine the degree of disease bur- chronic conditions and days of reduced activity; the pic- den and illness, and unhealthy risks can be passed on ture for women appeared less clear [18]. Lee et al found through families, communities, and populations follow- that low income males were more likely to report mental ing demographic gradients [4]. At different life stages, health problems, chronic symptoms and acute symptoms common risk factors and determinants include poor intra- than their high income counterparts [19]. Similar findings uterine conditions; stress, violence and traumatic experi- were reported in other studies, and risk factors associated ences; educational disadvantage; inadequate living envi- with chronic diseases also were also associated with low ronments that fail to promote healthy lifestyles; poor diet income [15,19-21]. and lack of exercise; alcohol misuse and tobacco smoking [5,30]. Risk factors are also increasingly more prevalent in The work of Mathers is significant for its systematic docu- areas of low socioeconomic status and in communities mentation of health inequalities among working aged characterised by low levels of educational attainment; Australians (25 to 64 years) in the late 1980s. He exam- high levels of unemployment; substantial levels of dis- ined mortality, disability, disease groups, specific dis- crimination, interpersonal violence and exclusion; and eases, self-perceived health, risk factors, health service use poverty. There is a higher prevalence of such factors and use of preventive screening services, using data from among Indigenous communities, and other socioeco- the 1989–90 National Health Survey (NHS) [15]. Mathers nomically disadvantaged Australians [5,6]. found that there were no clear gradients of chronic, recent or minor illness with level of socioeconomic disadvantage The inequalities in health observed across populations are of area, although there were some specific health status many – some of them are inevitable and others, unneces- indicators (self-reported health, reduced activity, unhap- sary and unfair. Those inequalities that are potentially piness) and certain risk factors (inactivity, smoking and avoidable are deemed 'inequitable' [7]. Despite signifi- alcohol use) that were reported more frequently by those cant medical advances and improved public health in in the more disadvantaged quintiles [15]. recent decades, socioeconomically disadvantaged com- munities continue to suffer an unequal burden of illness, Current information on inequalities in health other than premature death and disability. Therefore, the study of mortality is limited in Australia because of a dearth of socioeconomic inequalities in chronic diseases and condi- suitable data collections. However, the release of data tions and in risk factors is important and necessary. This is from the 2001 NHS provides an opportunity to examine particularly so, if we wish to develop more effective policy the prevalence of self-reported chronic disease in Australia mechanisms for preventing and intervening earlier in the and the way in which this impacts on different socioeco- progression of chronic diseases and their associated risk nomic groupings within the population. factors across the diverse Australian population, and to reduce some of the existing health inequities. Results Information for a selection of chronic diseases is shown in Our approach Table 1. Diseases were included on the basis of either high There have been a number of studies published in Aus- prevalence or their contribution to the burden of disease. tralia on socioeconomic inequalities in mortality from various chronic diseases and conditions. The earlier ones The main findings are: were analysed using information on occupation recorded on the death certificate [8-10]. An alternative approach The largest differential between those in the most well off has been to examine variations in mortality rates by and those in the most disadvantaged areas was for Page 2 of 8 (page number not for citation purposes) Australia and New Zealand Health Policy 2004, 1:8 http://www.anzhealthpolicy.com/content/1/1/8 Table 1: Inequality in prevalence of selected chronic diseases , 2001 2 3 Age group (years) and chronic disease Rate Rate ratio by quintile of socioeconomic disadvantage of area First Second Third Fourth Fifth 0–14 Mental and behavioural problems 6 596 1.00 1.04 1.10 1.12 1.52** Respiratory system 21 807 1.00 1.07 1.05 1.11 0.99 Asthma 13 363 1.00 1.10 1.12 1.25* 1.12 15–24 Mental and behavioural problems 10 284 1.00 1.02 0.97 1.08 1.28 Respiratory system 33 373 1.00 1.04 1.12 1.09 1.00 Asthma 16 263 1.00 0.82 1.14 1.02 1.00 6 6 6 6 6 Bronchitis/emphysema 1 701 1.00 1.32 1.66 1.94 1.97 Musculoskeletal system 19 088 1.00 1.11 1.00 1.08 0.94 25–64 Diabetes mellitus 2 234 1.00 1.37 1.67* 1.72* 2.28*** Mental and behavioural problems 11 093 1.00 1.05 1.20* 1.36*** 1.67*** Circulatory system 17 491 1.00 1.04 0.97 1.15* 1.28*** Hypertensive disease 9 751 1.00 1.12 1.01 1.24* 1.54*** Respiratory system 32 964 1.00 1.00 0.99 0.99 1.01 Asthma 10 393 1.00 1.10 0.99 1.19* 1.14 Bronchitis/emphysema 3 429 1.00 0.97 1.14 1.55** 1.70*** Digestive system 8 074 1.00 1.03 1.07 1.12 1.37*** Musculoskeletal system 39 840 1.00 1.10* 1.16*** 1.16*** 1.22*** 65 & over Diabetes mellitus 8 981 1.00 1.13 1.14 1.52* 1.56* Mental and behavioural problems 7 222 1.00 1.21 1.62* 1.67* 1.56* Circulatory system 56 592 1.00 1.09 1.06 1.10 1.19* Respiratory system 31 442 1.00 1.03 0.87 0.95 1.22* Musculoskeletal system 63 669 1.00 1.02 1.06 1.03 1.08 Survey respondents can report more than one disease. Rate is the number of persons per 100,000 population reporting the disease. The extent of any inequality is shown by the rate ratio, which expresses the ratio of the rate in each quintile to the rate in Quintile 1 (the most advantaged areas, with a rate ratio of 1.00); rate ratios differing significantly from 1.0 are shown with * p < 0.05; ** p < 0.01; *** p < 0.001. Information about these age groups were collected by proxy, using parental report. Information may be based on self-diagnosis, rather than diagnosis by a health practitioner. Indicates rate ratio based on estimates with a Relative Standard Error of between 25% and 50% and should be used with caution. Includes diseases of the connective tissue. Source: National Health Survey, ABS 2002 diabetes mellitus at ages 25 to 64 years, with the preva- Circulatory system diseases (in particular, hypertensive lence in the most disadvantaged areas being just over two disease) and digestive system diseases also exhibit a strong and a quarter times (a rate ratio of 2.28) the prevalence for differential in the 25 to 64 year age group (statistically sig- the least disadvantaged; there is also a strong, continuous nificant differentials of 28% and 54%, respectively). gradient in the rates, with the rate ratios in each of the third to fifth quintiles statistically significant. In the 65 years and over age group, the strongest differen- tials were evident for mental and behavioural problems (a There was a statistically significant differential of 67% at statistically significant 56%), diabetes (with a continuous ages 25 to 64 years, with a strong, continuous gradient, in gradient in rates, statistically significant in quintile3 four the prevalence of self-reported mental and behavioural and five) and respiratory system disease (a statistically sig- problems across the socioeconomic gradient; differentials nificant 22%). (also statistically significant) in the 0 to 14 year and 65 years and over age groups were 52% and 56%, Asthma accounted for almost two thirds of the rate of respectively. reporting of respiratory system disease in the 0 to 14 year Page 3 of 8 (page number not for citation purposes) Australia and New Zealand Health Policy 2004, 1:8 http://www.anzhealthpolicy.com/content/1/1/8 Table 2: Inequality in prevalence of selected health risk factors, 18–64 years, 2001 2 3 Health risk factors Rate Rate ratio by quintile of socioeconomic disadvantage of area First Second Third Fourth Fifth Current smokers - Male 30 582 1.00 1.40*** 1.55*** 1.71*** 1.95*** - Female 24 009 1.00 1.29*** 1.34*** 1.48*** 2.00*** - Persons 27 275 1.00 1.35*** 1.45*** 1.61*** 1.96*** Alcohol – High risk - Males 6 976 1.00 1.09 1.26* 1.26* 1.45*** - Females 2 127 1.00 0.59* 0.94 0.76 0.87 - Persons 4 537 1.00 0.93 1.16 1.12 1.22* Did not exercise - Males 28 772 1.00 1.20** 1.36*** 1.52*** 1.68*** - Females 28 220 1.00 1.19** 1.29*** 1.35*** 1.65*** - Persons 28 494 1.00 1.20*** 1.32*** 1.43*** 1.66*** Underweight females 12 675 1.00 0.89 0.83* 0.72*** 0.91 Overweight/obese - Males 54 701 1.00 1.09* 1.11* 1.04*** 1.00 - Females 37 004 1.00 1.09 1.21*** 1.16** 1.17** - Persons 45 798 1.00 1.09** 1.15*** 1.09** 1.06 Survey respondents can be shown under more than one type of risk factor. Rate is the number of persons per 100,000 population estimated to be at risk from the health risk factor. The extent of any inequality is shown by the rate ratio, which expresses the ratio of the rate in each quintile to the rate in Quintile 1 (the most advantaged areas, with a rate ratio of 1.00); rate ratios differing significantly from 1.0 are shown with * p < 0.05; ** p < 0.01; *** p < 0.001. Source: National Health Survey, ABS 2002 age group, almost half in the 15 to 24 year age group, and level measures of SES were used [28]. Thus, chronic dis- for about a third of the rate in the 25 to 64 year age group. ease inequalities in the wider population by SES are likely to be larger than those reported in this study. In addition, The NHS also included data on a number of risk factors the exclusion of the 'sparsely settled' areas of Australia in for chronic diseases, namely self-reported smoking, alco- NHS data collection results in the omission of data from hol misuse, physical inactivity and excess weight. A a high percentage of Indigenous people, who are the pop- number of these risk factors show a striking association ulation group with the poorest health. with socioeconomic status, in particular for people who are smokers and those who did not exercise, with contin- Discussion uous gradients and highly elevated rates of statistical sig- Our analysis indicates that socioeconomic inequalities in nificance (Table 2). The differences in male and female the prevalence of chronic diseases and their concomitant rates are also of interest. It was only for underweight risk factors are evident across the Australian population. females, and for the risk factor of high-risk alcohol con- However, the diseases with substantial disparities across sumption by females, that the socioeconomic gradient the socioeconomic quintiles are different, for different was reversed. stages in the life course. Although these results cannot be directly compared with those of previous studies, because It is important to note that the inequalities reported above of definitional and methodological differences, the recur- relate to the health of those people living in a geographic ring finding of inequalities for chronic disease morbidity area and to the overall level of socioeconomic disadvan- and risk factor prevalence across the socioeconomic gradi- tage of that area. Most areas will contain varying levels of ent remains a significant concern. individual socioeconomic disadvantage and, to the extent that the poorer health is associated with individual The burden in the Australian population attributable to economic circumstances and living conditions rather than socioeconomic inequality is large, and has far-reaching communal environment, the inequalities will understate implications in terms of unnecessary disability and suffer- the true differences in health status according to socioeco- ing, the loss of potentially economically productive mem- nomic disadvantage [15]. bers of society, and increased costs for the health and social care systems [35]. Despite the expenditure of mil- Furthermore, there are limitations to the use of area-based lions of dollars to prevent and reduce the prevalence of measures of SES. Due to misclassification error (i.e. ascrib- chronic diseases and their risk factors, these inequities ing area-SES to individuals), estimates of difference across have persisted. However, the situation in Australia is by the quintiles will be smaller than if data on individual- no means unique, for inequalities in these diseases and Page 4 of 8 (page number not for citation purposes) Australia and New Zealand Health Policy 2004, 1:8 http://www.anzhealthpolicy.com/content/1/1/8 their risk factors have been observed for most of the devel- Policy-makers who wish to address socioeconomic ine- oped countries in which they have been studied [26]. qualities in health may favour one of the following approaches. Some view the impact of socioeconomic dis- What should we be doing differently? There is a growing advantage on those groups with the poorest health in the body of knowledge that will help to provide direction for population, such as Aboriginal people and Torres Strait developing policies to reduce inequities across the popu- Islanders, as the priority policy goal. Others identify the lation. The socioeconomic environment is a powerful and gap between the health of those groups at the outer ends potentially modifiable factor, and public policy is a key of the socioeconomic hierarchy (those with the poorest instrument to improve this environment, particularly in health and those with best health), and see the narrowing areas such as housing, taxation and social security, work of the gap as the goal. Others prefer to focus on the socio- environments, urban design, pollution control, educa- economic gradient in health that runs across the whole tional achievement, and early childhood development population [31]. [34]. Graham has identified that the last approach widens the However, attention must be paid to the nature of any policy debate in three ways [31]. Firstly, it looks for the action that is taken, to ensure that social and economic causes of health inequality in the systematic differences in inequalities are not increased. Some programs, by their life chances and opportunities, living standards and life- very success, can increase inequality by widening the gap styles that are associated with people's unequal positions between groups in the population; for example, such pro- across the socioeconomic hierarchy, and for the pathways grams may be more attractive to those who are already through which they influence health [31,36]. Secondly, as healthier, or not as effective for certain groups with poorer a result, addressing health inequalities becomes a popula- health, less education or more stressful lives. In one smok- tion-wide goal that includes every citizen [31]. Thirdly, ing cessation initiative, it was found that the prevalence of 'reducing health gradients' provides a more comprehen- smoking decreased predominately in those adults with sive policy approach: one that encompasses 'remedying higher education, thus increasing the existing difference disadvantages' and 'narrowing health gaps' within the with those who were more disadvantaged [37]. While broader goal of 'equalising health chances across all the smoking prevalence in Australia has reduced considerably socioeconomic groups' [31]. over the last 20 years, attributes such as lower education and occupational status, unemployment, rented housing, She also observes that, "improving the health of poor and living in disadvantaged areas reflect a higher proba- groups and improving their position relative to other bility of reporting tobacco expenditure [32]. As a result, groups are necessary elements in a strategy to reduce the the tax revenue from the sale of tobacco products is being socioeconomic gradient. However, neither is sufficient on disproportionately drawn from the poorest households its own. To reduce the socioeconomic gradient, health in and represents a greater proportion of their household other socioeconomic groups also needs to improve at a budget [32]. faster rate than in the highest socioeconomic group. Thus, policies to ameliorate health disadvantage, to close health It is also evident that the ways in which systems such as gaps and to reduce health gradients need to be pursued education and health are delivered and structured can together, and not at the expense of each other" [31]. increase existing inequality. For example, schooling can be a way of addressing inequality and also a way of repro- There is also an urgent need to make health inequalities a ducing it. It has been suggested that there are two goals for research priority for each stage of the life course – not just a social justice program in education: to work to eliminate to monitor the size and extent of the disparities but also the contribution that the education system makes to the to undertake research that will find preventive approaches production over time of social inequality in general; and and further policy interventions that will be effective in to maximise the positive contributions that the education reducing them, and that are likely to be implemented by system makes to reducing social inequality [33]. There- governments and communities. fore, different approaches and mixes of policies and pro- grams must be mounted to address inequalities. These Conclusions approaches may include more precise targeting, but also Clearly, any moves to address the impact of chronic dis- greater attention to community-based dimensions of ease at a population level must take into account socioe- 'interdependence' between individual behaviours, key conomic inequalities in prevalence. More research is determinants, and community and institutional needed to determine which approaches are effective and resources. why others have failed to have the desired impact, partic- ularly for those who are from socioeconomically disad- vantaged areas. Finally, although rates are generally Page 5 of 8 (page number not for citation purposes) Australia and New Zealand Health Policy 2004, 1:8 http://www.anzhealthpolicy.com/content/1/1/8 highest at the oldest ages, the development of risk factors the classification developed by Mathers [15]; and for for many chronic diseases occurs early in life, and thus, it which sufficient cases were available for analysis by five is essential those health inequities are addressed right groups of socioeconomic disadvantage of area (see below across the life course. for details of the way these groups were constructed). Methods The risk factors used by the ABS were those identified for Data sources the NHPA conditions [27]. The ABS conducts the National Health Survey (NHS) on a regular basis, most recently in 2001 [22]. The NHS collects The ABS has coded conditions reported by respondents to information from approximately 26,900 people from all output disease categories based on ICD-10. Conditions States and Territories living in private dwellings, selected described as 'chronic' in this article include those long- at random using a multi-stage area sample of private term conditions reported in the NHS, which are com- dwellings. The survey is undertaken across much of Aus- monly recognised by health practitioners as chronic dis- tralia, but excludes the 'sparsely settled' areas, which com- eases [23]. The risk factor for 'high risk due to alcohol' prised less than 1% of the non-Indigenous population reflects the National Health and Medical Research and 25% of the Indigenous population at the 2001 Cen- Council's risk levels for harm in the long term from alco- sus: a separate Australia-wide survey of the health of hol consumption [24]. The risk factors for overweight and Indigenous people, also conducted in 2001, surveyed underweight were calculated from self-reported height these sparsely settled areas. and weight information and grouped to reflect World Health Organization (WHO) guidelines [25]. The survey includes self-reported details of health condi- tions (both acute and long term) and major risk factors, as Given the policy importance of the NHPAs, the 2001 NHS well as demographic and socioeconomic information questionnaire underwent significant revisions to more about the survey respondent. Respondents were asked if precisely capture information on several of the NHPAs. they had been told by a doctor or nurse that they had Consequently, while the quality of the information on asthma, cancer, heart and circulatory conditions, and/or NHPAs has been improved from the 1995 NHS, the diabetes. These conditions, together with injuries and degree of comparability with previous surveys has been mental health, form the NHPAs [27]. However, for long somewhat compromised for many of the major health term mental health problems, respondents were not asked conditions. Some specific conditions (e.g., diabetes) whether they had been told by a doctor or nurse that they appear to be comparable between the 1989–90 and 2001 had any mental health problems; thus, the responses may surveys, however, for most groups of conditions based on be based on self-diagnosis, rather than diagnosis by a ICD chapter headings (e.g., all circulatory) the ABS advise health practitioner [22]. Respondents were also asked a that the combined effect of major conceptual changes as series of questions about other specific, non-NHPA, con- well as major classification changes between the 1989–90 ditions, covering eye and sight problems, ear and hearing and 2001 surveys would make direct comparisons very problems, and arthritis, rheumatism and gout. They were difficult. This analysis is therefore restricted to the 2001 then shown a series of three prompt cards (two with con- data. ditions listed, while the third contained more general Measurement of socioeconomic status descriptions of condition types) and asked whether they had any of the conditions shown or conditions similar to The socioeconomic status (SES) of the address of resi- those shown or described. In each of these cases, details dence of each survey respondent is available at the Census were recorded for conditions reported as current at the Collection District (CD) level and was added to the NHS time of the survey; respondents were also asked whether file, as was the quintile of socioeconomic disadvantage of the condition had lasted, or was expected to last, for six area into which that CD fell at the Census. The measure months or more. Information was gathered directly from used to allocate CDs to quintiles was the 1996 Census individuals aged 15 years and older. For children up to the Index of Relative Socio-Economic Disadvantage (IRSD). age of 15 years, information was provided by proxy, from The IRSD is one of five Socio-Economic Indexes for Areas a parent or guardian. produced by the ABS using Principal Components Analy- sis. It summarises information available from variables The particular conditions for which data were requested collected in the 1996 Population Census including those from the ABS for this analysis were: related to education, occupation, and income. The varia- bles are expressed as percentages of the relevant the NHPA conditions (other than injury, which has not population. been included in this paper, with its focus on chronic dis- ease); plus other 'serious' chronic conditions, in line with Page 6 of 8 (page number not for citation purposes) Australia and New Zealand Health Policy 2004, 1:8 http://www.anzhealthpolicy.com/content/1/1/8 The NHS records were aggregated to the quintiles derived ratio of the rate in each quintile to the rate in Quintile 1 from the Census data, where Quintile 1 comprises the (the most advantaged areas, with a rate ratio of 1.00). CDs with the highest IRSD scores (highest socioeconomic status, or most advantaged, areas) and Quintile 5 com- Competing interests prises the CDs with the lowest IRSD scores (lowest socio- The author(s) declare that they have no competing economic status, or most disadvantaged areas). Each interests. quintile comprises approximately 20% of CDs. Authors' contributions The ABS produced the estimates of the number of people JG conceived the study, and was responsible for its design with chronic diseases and risk factors by quintile. The and coordination. ST participated in the design of the method used resulted in the production of quintiles of study and performed the statistical analysis. DH partici- varying sizes, ranging from 17.4% of the population in pated in the drafting of the manuscript. All authors read Quintile 5 (most disadvantaged areas) to 22.8% in Quin- and approved the final manuscript. tile 2 and 21.1% in Quintile 1 (most advantaged areas). This is a differential of over five percentage points Acknowledgements The Australian Government Department of Health and Ageing funded this between Quintile 5 and Quintile 2, or 1,027,030 fewer research. The data were provided by the ABS. people in the most disadvantaged areas when compared with Quintile 2 (and 708,980 fewer in Quintile 5 than in References Quintile 1). The effect of this lack of precision on the 1. Mathers CD, Vos ET, Stevenson CE, Begg SJ: The Australian Bur- results by quintile is not known. Although, in part, the dif- den of Disease Study: measuring the loss of health from dis- ference arises as a result of the method used (that is, that eases, injuries and risk factors. Med J Aust 2000, 172:592-596. 2. Australian Institute of Health and Welfare: Chronic diseases and associ- the quintiles are based on the Census population, and ated risk factors in Australia, 2001 Canberra; 2002. [AIHW Catalogue applied to the NHS population which has a different pro- No. PHE 33] 3. Thacker S, Stroup D, Rothenberg R: Public health surveillance for file), it may also reflect different response rates to the sur- chronic conditions: a scientific basis for decisions. Statistics in vey from different socioeconomic groups. Medicine 1995, 14:629-641. 4. Ackland M, Choi BC, Puska P: Rethinking the terms non-com- municable disease and chronic disease. J Epidemiol Community The NHS includes other measures that could potentially Health 2003, 57:838-9. be used to measure socioeconomic status. For example, 5. National Public Health Partnership (NPHP): Preventing Chronic Disease: A Strategic Framework Melbourne, Victoria; 2001. income reporting in ABS surveys is known to be incom- 6. Mooney G: Inequity in Australian health care: how do we plete, in particular for low income groups and, in the progress from here? Aust NZ J Pub Health 2003, 27:267-270. 2001 NHS, income was only collected for the respondent 7. Whitehead M: The concepts and principles of equity in health. Int J Health Serv 1992, 22:429-445. and their spouse. Information on education is also avail- 8. McMichael AJ, Hartshorne JM: Mortality risks in Australian men able, but a single measure has yet to be agreed. Problems by occupational groups, 1968–1978: variations associated in using education include that it is typically completed with differences in drinking and smoking habits. Med J Aust 1982, 1:253-6. early in adulthood; it captures neither differential on-the- 9. Taylor R, Herman H, Preston G: Occupation and mortality in Australian job training and other career investments made by indi- working age males, 1975–77 Health Commission of Victoria and Department of Social and Preventive Medicine, Monash University: viduals with similar levels of formal schooling, nor the Melbourne; 1983. volatility in economic status during adulthood that has 10. McMichael AJ: Social class (as estimated by occupational pres- been shown to have adverse implications for health [29]. tige) and mortality in Australian males in the 1970s. Commu- nity Health Stud 1985, 9:20-30. The age structure of the population may also influence the 11. Siskind V, Najman J, Copeman R: Socioeconomic status and mor- indicator: an older population generally has lower educa- tality: a Brisbane area analysis. Community Health Stud 1987, 11:15-23. tion levels than a younger population due to improved 12. South Australian Health Commission (SAHC): A Social Health Atlas of access to education over time. South Australia. Adelaide 1990. 13. McNeil J, King R, Jennings G, Powles J, eds: A textbook of preventative medicine Melbourne: Edward Arnold; 1990. Analysis 14. National Health Strategy: Enough to make you sick: how income Chronic disease and risk factor rates are expressed as rates and environment affect health. [National Health Strategy Research per 100,000 population, indirectly age-standardised. The Paper No.1]. Melbourne 1992. 15. Mathers C: Health inequalities among adult Australians aged rates were also calculated by age although, given the small 25–64 years. In [Australian Institute of Health and Welfare Health Mon- sample size of the NHS, only broad age groups (0–14 itoring Series No. 1] Canberra: AGPS; 1994. 16. Glover J, Harris K, Tennant S: A Social Health Atlas of Australia 2nd edi- years, 15–24 years, 25–64 years and 65 years and over) tion. Adelaide: Openbook; 1999. were available. The standard population and quintile 17. Turrell G, Mathers C: Socioeconomic inequalities in all-cause populations are the weighted 2001 survey populations and specific-cause mortality in Australia: 1985–1987 and 1995–1997. Int J Epidemiol 2001, 30:231-239. from the NHS. The extent of any differential between the 18. Broadhead P: Social status and morbidity in Australia. Commu- quintiles is shown by the rate ratio, which expresses the nity Health Stud 1985, 9:87-98. 19. Lee S-H, Smith L, d'Espaignet E, Thompson N: Health Differentials for Working Age Australians Canberra: Australian Institute of Health; 1987. Page 7 of 8 (page number not for citation purposes) Australia and New Zealand Health Policy 2004, 1:8 http://www.anzhealthpolicy.com/content/1/1/8 20. Broom D: The social distribution of illness: Is Australia more equal? Soc Sci Med 1984, 18:909-917. 21. Dobson A, Gibberd R, Leeder S, O'Donnell D: Occupational dif- ferences in ischaemic heart disease mortality and risk fac- tors in Australia. Am J Epidemiol 1985, 122:282-90. 22. Australian Bureau of Statistics (ABS): National Health Survey: Summary of Results Canberra: ABS; 2002. [ABS Cat. No. 4364.0] 23. Rothenberg R, Koplan J: Chronic disease in the 1990s. Annu Rev Public Health 1990, 11:267-96. 24. National Health and Medical Research Council (NHMRC): Australian Alcohol Guidelines: Health Risks and Benefits Canberra: AusInfo; 2001. 25. World Health Organization (WHO): Obesity: Preventing and Managing the Global Epidemic – Report of a WHO Consultation. Geneva 2000. [WHO Technical Report Series, No. 894] 26. Beaglehole R, Yach D: Globalisation and the prevention and control of non-communicable disease: the neglected chronic diseases of adults. Lancet 2003, 362:903-08. 27. Australian Government Department of Health and Ageing: National Health Priority Areas. [http://www.health.gov.au/pq/nhpa/]. accessed 27 July 2004 28. Hyndman JC, Holman CSD, Hockey RL, Donovan RJ, Corti B, Rivera J: Misclassification of social disadvantage based on geograph- ical areas: comparison of postcode and collector's district analyses. International Journal of Epidemiology 1995, 24:165-176. 29. McDonough P, Duncan DJ, Williams DR, House J: Income dynam- ics and adult mortality in the US, 1972–1989. American Journal of Public Health 1997, 18:341-378. 30. Williams DR: Race, socioeconomic status, and health: the added effects of racism and discrimination. In In: Socioeconomic status and health in industrial nations: social, psychological, and biological pathways Edited by: Adler NE, Marmot M, McEwen S, Stewart J. New York: Annals of the New York Academy of Science; 1999:173-188. 31. Graham H: Tackling Inequalities in Health in England: Reme- dying Health Disadvantages, Narrowing Health Gaps or Reducing Health Gradients? Journal of Social Policy 2004, 33(1):115-131. 32. Siahpush M, Borland R, Scollo M: Smoking, socioeconomic status, and household expenditure on food, alcohol, gambling and insurance The Can- cer Council of Victoria: Carlton South; 2003. 33. Connell RW, White V, Johnston K, eds: Running Twice As Hard: the dis- advantaged schools program in Australia Geelong: Deakin University; 34. Hetzel D, Page A, Glover J, Tennant S: Inequality in South Australia: The Evidence Volume 1. Adelaide: Department of Health, SA; 2004. 35. Centre for South Australian Economic Studies (CSAES): The Impact of Socio-Economic and Locational Disadvantage on Health Outcomes and Cost. In Social Justice Research Program into Locational Disadvantage – Report no. 10 Department of Health, Housing, Local Government and Community Services: AGPS, Canberra; 1993. 36. Davey Smith G, Gunnell D, Ben-Shlomo Y: Life-course approaches to socioeconomic differentials in cause-specific adult mortal- ity'. In in Poverty, Inequality and Health Edited by: Leon D, Walt G. Oxford: Oxford University Press; 2001. 37. Osler M, Gerdes LU, Davidsen M, Brønnum-Hansen H, Madsen M, Jørgensen T, Schroll M: Socioeconomic status and trends in risk factors for cardiovascular diseases in the Danish MONICA population, 1982–1992. Journal of Epidemiology and Community Health 2000, 54:108-13. Publish with Bio Med Central and every scientist can read your work free of charge "BioMed Central will be the most significant development for disseminating the results of biomedical researc h in our lifetime." Sir Paul Nurse, Cancer Research UK Your research papers will be: available free of charge to the entire biomedical community peer reviewed and published immediately upon acceptance cited in PubMed and archived on PubMed Central yours — you keep the copyright BioMedcentral Submit your manuscript here: http://www.biomedcentral.com/info/publishing_adv.asp Page 8 of 8 (page number not for citation purposes)

Journal

Australia and New Zealand Health Policy – Springer Journals

Published: Dec 12, 2004

There are no references for this article.