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

Learn More →

Socio‐economic gradients in self‐reported diabetes for Indigenous and non‐Indigenous Australians aged 18–64

Socio‐economic gradients in self‐reported diabetes for Indigenous and non‐Indigenous Australians... D iabetes is an important cause of morbidity and mortality worldwide. In developed countries, diabetes is more common among those of lower socio‐economic status, but the reverse has been true in some developing countries. The burden of diabetes is particularly pronounced among Indigenous Australians, but little is known about its distribution within the Indigenous population. A recent analysis found significant inverse socio‐economic gradients in diabetes prevalence among urban Indigenous participants in a study in northern Australia, but it is not clear whether this finding is indicative of trends in the Indigenous population elsewhere, or whether any socio‐economic gradients in the Indigenous population are of similar magnitude to those in the non‐Indigenous population. The aim of the current study is to examine socio‐economic gradients in diabetes among a nationally representative sample of Indigenous Australians, and to compare these with corresponding gradients in the non‐Indigenous population. Methods Data for Indigenous and non‐Indigenous adults aged 18–64 years were taken from two national surveys conducted in parallel by the Australian Bureau of Statistics (ABS) in 2004–05: the National Aboriginal and Torres Strait Islander Health Survey (NATSIHS) and the National Health Survey (NHS). These two surveys had very similar content and in most cases the wording of questions on particular topics was identical. This analysis is limited to responses to questions deemed by the ABS to be comparable in the two surveys. The NATSIHS was conducted using two different multi‐stage sampling strategies. In discrete Indigenous communities in remote areas in the Northern Territory, Queensland, South Australia and Western Australia, communities were randomly selected, with probability of selection proportional to community size. A random sample of dwellings within these communities was selected, with one Indigenous adult and up to one Indigenous child then randomly selected to participate. In the remainder of Australia, Census collection districts (CDs) were stratified by state, remoteness classification and the number of Indigenous dwellings in the 2001 Census. A sample of CDs was selected, with probability of selection based on the number of Indigenous households in 2001. Within selected CDs, dwellings were randomly selected and screened to determine whether they included Indigenous members. If they did, then up to two Indigenous adults and up to two Indigenous children were randomly selected to participate. In addition, Indigenous respondents from the NHS were included with NATSIHS data to provide Indigenous population estimates. The NHS was conducted during the same timeframe as the NATSIHS using similar methods and materials. Dwellings were selected using a multi‐stage sampling strategy, and one adult and up to one child were randomly chosen within selected dwellings. Very remote areas were out of scope in the NHS. Both surveys were limited to usual residents of private dwellings, and both were conducted by trained ABS interviewers. In both the NHS and in non‐remote areas in the NATSIHS, data were collected using a Computer Assisted Interview technique. In remote areas of the NATSIHS, pen and paper interview forms were used and some questions were simplified or deleted. Adults aged 18 years and over were personally interviewed. More details about the design and conduct of the surveys have been published elsewhere. To allow data access to interested researchers, the ABS created a Confidentialised Unit Record File (CURF) for the NATSIHS. This file includes unit records for Indigenous respondents of the 2004–05 NATSIHS and the 2004–05 NHS, as well as unit records for non‐Indigenous respondents from the 2004–05 NHS. Although the CURF contains unit records for participants of all ages, this analysis is limited to data from the 20,849 respondents (5,417 Indigenous and 15,432 non‐Indigenous) aged 18–64 years. The exclusion of those aged 65 years and older was due to uncertainty about the applicability of socio‐economic indicators among older people, as well as the relatively small size of this group in the Indigenous population. Definition of diabetes Diabetes was considered to be present if the respondent indicated he or she currently had diabetes or high sugar levels. Diabetes that was reported as not current was not included. All current diabetes was assumed in the surveys to be long‐term (that is, having lasted, or expected to last, at least six months). Almost all respondents who reported diabetes or high sugar levels (referred to hereafter as diabetes) reported that they had been told this by a health care provider. No information was available about type of diabetes, and it was not possible to distinguish between diabetes and high sugar levels as these were not asked about separately in remote areas. Socio‐demographic factors Information was available on a range of socio‐economic and demographic factors, including age, sex, educational attainment, non‐school qualifications, employment status, household income, home ownership (Indigenous respondents only), area of residence and area‐level disadvantage. Available categories for these variables are as shown in Table 1 , although these have been combined for analysis as appropriate. Information about age, sex and whether the respondent was currently attending school was provided by ‘any responsible adult’ within the household; information about the dwelling (including tenure) and the income of non‐participant household members (required to calculate household income) was provided by a household ‘spokesperson’, chosen on the basis of his or her ability to provide accurate information. Information relating to geography (including remoteness classification and area‐level disadvantage score) was provided by the ABS based on the CD in which the selected dwelling was located. All other information used in this analysis was provided by the respondent. 1 Socio‐demographic characteristics of Indigenous and non‐Indigenous adults aged 18–64. a Indigenous % (95% CI) b Non‐Indigenous % (95% CI) b Age (years) 18–24 23.1 (21.7–24.4) 15.1 (14.8–15.4) 25–34 28.4 (27.7–29.0) 22.4 (22.3–22.6) 35–44 24.0 (23.5–24.5) 23.5 (23.4–23.7) 45–54 16.1 (15.7–16.4) 22.0 (21.8–22.1) 55–64 8.5 (7.1–9.9) 17.0 (16.9–17.1) Male 46.8 (45.6–47.9) 49.8 (49.6–50.1) Highest year of school completed Year 12 23.5 (21.2–25.8) 52.5 (51.2–53.8) Year 11 13.0 (11.7–14.4) 10.9 (10.3–11.6) Year 10 31.2 (29.4–33.1) 24.7 (23.7–25.7) Year 9 13.9 (12.5–15.3) 6.3 (5.8–6.7) Year 8 or less 17.3 (15.7–18.9) 5.4 (4.8–5.9) Never went to school 1.1 (0.5–1.6) 0.2 (0.1–0.3) Level of highest non‐school qualification Post‐graduate degree 1.9 (1.0–2.7) 6.2 (5.8–6.7) Bachelor's degree 2.9 (2.3–3.6) 14.5 (13.8–15.3) Diploma 4.7 (3.7–5.7) 9.7 (9.1–10.3) Certificate 24.2 (22.2–26.1) 26.0 (25.0–27.1) No qualifications 66.4 (64.1–68.6) 43.5 (42.5–44.5) Employment status Employed full‐time 32.7 (30.4–35.1) 54.3 (53.4–55.2) Employed part‐time 21.9 (20.1–23.8) 21.8 (21.1–22.5) Unemployed 8.1 (6.9–9.2) 3.0 (2.7–3.4) Not in the labour force 37.3 (35.0–39.6) 20.9 (20.1–21.7) Housing tenure Owner‐occupied without mortgage 6.8 (5.4–8.3) n/a c Owner‐occupied with mortgage 17.9 (15.7–20.0) n/a c Renter or other tenure 75.3 (72.7–77.9) n/a c Equivalised household income quintile d 1 (lowest) 33.7 (31.4–36.1) 11.3 (10.7–11.9) 2 21.6 (19.7–23.6) 13.1 (12.5–13.8) 3 14.3 (12.4–16.1) 16.9 (16.1–17.6) 4 9.4 (7.7–11.2) 19.5 (18.7–20.2) 5 (highest) 5.2 (4.0–6.4) 21.7 (20.7–22.7) Not known or not stated 15.6 (13.6–17.6) 17.5 (16.6–18.4) SEIFA quintile e 1 (most disadvantaged) 49.3 (43.7–55.0) 17.1 (15.7–18.5) 2 19.3 (15.2–23.3) 19.0 (17.4–20.7) 3 18.5 (14.3–22.7) 20.3 (18.4–22.2) 4 9.0 (6.4–11.6) 21.3 (19.5–23.0) 5 (least disadvantaged) 3.9 (2.2–5.7) 22.3 (20.0–24.7) Area of residence Major cities 30.6 (29.1–32.0) 70.2 (68.6–71.8) Inner regional 20.1 (19.0–21.3) 19.5 (17.9–21.0) Outer regional 21.5 (20.4–22.5) 10.4 (9.2–11.5) Remote or very remote 27.8 (26.3–29.4) — Notes: a) Source: Weighted data from the National Aboriginal and Torres Strait Islander Health Survey 2004–05 confidentialised unit record file (CURF). b) CI, confidence interval. Proportions are weighted to provide population estimates. Totals are based on those with non‐missing data, except for equivalised household income, for which a separate category is shown. c) n/a, not available. d) Gross weekly equivalised cash income of household, using the OECD scale. Quintiles are based on national figures. e) SEIFA, Socioeconomic Index for Areas, Index of Relative Disadvantage. Those reported to be still at school (n=67) were not asked questions about the highest year of school completed or non‐school qualifications. They have been coded as missing on both of these variables. Those whose educational attainment was not stated (n=2) and those whose level of qualifications could not be determined (n=222) were coded as missing on the relevant variable only. Gross weekly household equivalised income, which takes into account household size and composition, was estimated using the OECD scale. Quintiles were determined based on all‐Australian figures. That is, the same categories were used for both Indigenous and non‐Indigenous participants. Equivalised income quintile was not available for 2,941 respondents (14.1%). Analyses were conducted with these respondents coded as missing, as well as with them included using a special category of household income unknown. Home ownership was available in the CURF only for Indigenous respondents (missing for n=41), and was based on whether the home was owned or being purchased by any of its occupants (not necessarily the respondent). Area of residence was classified according to the Australian Standard Geographical Classification remoteness classification (based on the ARIA+ index) into major cities, inner regional, outer regional, remote and very remote. ABS documentation indicates that the remote/very remote category was to be used for Indigenous respondents only. Therefore, area of residence was re‐coded to missing for 312 non‐Indigenous respondents (2%) whose residence was listed as remote/very remote. Area‐level disadvantage was based on the 2001 socio‐economic Indexes for Areas (SEIFA) Index of Disadvantage score for the CD of the selected dwelling. Quintiles were determined based on all‐Australian figures. That is, the same categories were used for both Indigenous and non‐Indigenous respondents. Those with SEIFA quintile not known (n=313) were coded as missing. Statistical analysis All analyses were conducted using STATA version 10.0 via the ABS's Remote Access Data Laboratory (RADL). Under the RADL system, analysts submit statistical code to the ABS; the code is then run and the output made available to the analyst through a password‐protected web account. Analysts do not have direct access to unit record data at any time, and there are limits placed on the commands and outputs that are allowed, in order to protect the security and confidentiality of the data. All analyses used ABS‐generated person‐weights (or expansion factors) to adjust for disproportionate sampling of some groups. The estimates produced in this manner apply to the population as a whole, and not just the sample. Standard errors and 95% confidence intervals (CI) were calculated using replicate weights produced by the ABS using the Jackknife method (250 replicate weights for Indigenous respondents, 60 for non‐Indigenous respondents). These replicate weights allow the estimation of standard errors taking into account the complex design and weighting procedures used in the surveys. Although STATA version 10 incorporates a suite of procedures to analyse complex survey data, these commands are not allowed in the RADL system (Therese Lalor, ABS, personal communication, May 2009). Instead, commands from the svr module written by Nick Winter (available using the STATA command: search svr, net) were used. STATA code is available from the author on request. Directly age‐standardised estimates and 95% CIs were calculated using an alternative set of person‐weights and replicate weights produced by the ABS for that purpose. The standard population was the total Australian population as at 30 June 2001. Logistic regression was conducted separately for Indigenous and non‐Indigenous groups due to the different numbers of replicate weights for the two groups. All models were adjusted for age group and sex, with socio‐economic variables assessed individually and in combination. Household income and SEIFA score were both modeled using all five quintiles as well as comparing quintiles one and two with quintiles 3–5 combined, but only the latter results are presented. There were relatively few Indigenous participants in the top quintiles of these variables ( Table 1 ), and preliminary analysis indicated a similar prevalence of diabetes in the top three quintiles on both measures, allowing them to be combined with minimal loss of information. Participants with missing data were excluded only from analyses involving the variable for which they were missing data. Ethics approval This study was approved by the Human Research Ethics Committee of the Northern Territory Department of Health and Families and the Menzies School of Health Research. The proposal was considered and approved by both the Aboriginal sub‐committee and the main committee. Results Almost one in 10 Indigenous people (9.7%, 95% CI 8.6%‐10.8%) aged 18–64 reported they currently had diabetes or high blood sugar levels. This was much higher than for the non‐Indigenous population (3.0%, 95% CI 2.6–3.3), even before adjusting for the younger age distribution of the Indigenous population. Diabetes increased more sharply with age in the Indigenous population than in the non‐Indigenous population ( Figure 1 ). There was a 30‐fold increase in prevalence from the youngest to the oldest age group among Indigenous people, about twice that seen in the non‐Indigenous population. 1 Prevalence of self‐reported diabetes by age and Indigenous status, 2004/05. The socio‐demographic profile of the Indigenous population was significantly different from that of the non‐Indigenous population ( Table 1 ), with a younger age distribution, lower educational attainment, and greater levels of disadvantage across a range of indicators. After adjusting for age, diabetes prevalence was significantly higher among those of lower SES in both the Indigenous and non‐Indigenous populations ( Figures 2a–2g , Table 2 ). Area of residence was significantly associated with diabetes prevalence in the Indigenous population, but not in the non‐Indigenous population. The relative differences between low and high SES groups, as measured by age‐ and sex‐adjusted odds ratios, were generally of similar magnitude for the Indigenous and non‐Indigenous populations ( Table 2 ). For example, the relative odds of diabetes for the lowest quintile of equivalised household income (compared with quintiles 3–5 combined) was 2.3 (95% CI 1.6–3.4) for the Indigenous population and 2.0 (95% CI 1.5–2.8) for the non‐Indigenous population. The relative odds of diabetes for those who did not complete at least Year 10 (compared with those who did) was 1.8 for both the Indigenous (95% CI 1.4–2.3) and non‐Indigenous (95% CI 1.4–2.4) populations. Despite these similarities in relative terms, there were marked disparities in absolute terms, with Indigenous people of high SES having a greater prevalence of diabetes than non‐Indigenous people of low SES on every SES measure examined ( Figures 2a–2g ). 2a–c Age‐standardised prevalence of self‐reported diabetes by indicators of socio‐economic status, Indigenous and non‐Indigenous adults aged 18–64, 2004/05. 2d–g Age‐standardised prevalence of self‐reported diabetes by indicators of socio‐economic status, Indigenous and non‐Indigenous adults aged 18–64, 2004/05. 2 Relative odds of diabetes by socioeconomic status for Indigenous and non‐Indigenous adults aged 18–64. a Indigenous Adjusted OR (95% CI) b Non‐Indigenous Adjusted OR (95% CI) b Highest year of school completed Year 10 or more 1.0 1.0 Less than Year 10 c 1.8 (1.4–2.3) 1.8 (1.4–2.4) Level of highest non‐school qualification Bachelor's or post‐graduate degree 0.4 (0.2–0.9) 0.8 (0.6–1.1) Diploma 0.5 (0.2–1.3) 0.5 (0.3–0.8) Certificate 0.6 (0.4–0.8) 0.9 (0.7–1.2) No qualifications 1.0 1.0 Employment status Employed 1.0 1.0 Unemployed 1.9 (1.0–3.4) 0.9 (0.3–2.1) Not in the labour force 1.6 (1.2–2.2) 1.9 (1.4–2.6) Housing tenure Owner‐purchaser 0.4 (0.2–0.5) — Renter or other tenure 1.0 — Equivalised household income quintile d 1 (lowest) 2.3 (1.6–3.4) 2.0 (1.5–2.8) 2 1.7 (1.0–2.7) 1.5 (1.1–2.1) 3–5 (highest) 1.0 1.0 Not known or not stated 2.2 (1.3–3.6) 1.0 (0.7–1.4) SEIFA quintile e 1 (most disadvantaged) 1.6 (1.0–2.5) 2.1 (1.6–2.9) 2 0.9 (0.6–1.5) 1.5 (1.0–2.1) 3–5 (least disadvantaged) 1.0 1.0 Area of residence Major cities 1.0 1.0 Inner regional 1.2 (0.7–1.9) 1.1 (0.8–1.4) Outer regional 1.6 (1.0–2.3) 1.0 (0.8–1.3) Remote or very remote 2.5 (1.8–3.6) — Notes: a) Source: Weighted data from the National Aboriginal and Torres Strait Islander Health Survey 2004–05 confidentialised unit record file (CURF). b) OR, odds ratio. CI, confidence interval. All odds ratios are from logistic regression models adjusted for age group, sex and the individual variable listed. c) Includes those who never went to school. d) Gross weekly equivalised cash income of household, using the OECD scale. Quintiles are based on national figures. e) SEIFA, Socioeconomic Index for Areas, Index of Relative Disadvantage. Adjusting for multiple SES variables simultaneously resulted in some attenuation of the odds ratios, but the directions of the associations remained the same (data not shown). Among Indigenous people, home ownership retained a strong and significant inverse association with diabetes, even after adjustment for other SES variables (OR 0.4, 95% CI 0.3–0.7). Discussion The results presented here indicate that socio‐economic disadvantage explains some but not all of the higher prevalence of diabetes among Indigenous Australians. There are marked SES‐related gradients in self‐reported diabetes among both Indigenous and non‐Indigenous Australians, and Indigenous Australians are over‐represented in lower SES categories. However, the prevalence of diabetes is higher for Indigenous people than non‐Indigenous people in every SES group, often by a considerable margin. Indeed, for most variables considered, Indigenous Australians of high SES are more likely to report diabetes than non‐Indigenous Australians of low SES. The combination of a socio‐economic gradient within the Indigenous population and a gap between the Indigenous and non‐Indigenous populations suggests that traditional risk factors may not be sufficient to explain completely the patterns of diabetes among Indigenous Australians. Other factors that may operate across the socio‐economic spectrum, including racism and discrimination, stress, and a legacy of grief, loss and dispossession, may also play a role through a range of neuroendocrine, autonomic, metabolic, immune and/or behavioural pathways. A recent study indicated that racial disparities in the US may be explained by differences in the ‘health risk’ environments in which African Americans and whites live, and this could be relevant in Australia. Differential access to services may also exist across SES groups. Although genes clearly play a role in the development of diabetes, the relationship between ethnicity and genetic susceptibility is quite complex. These data confirm the relationships previously observed in the DRUID Study among Indigenous people in one northern Australian city, in which home ownership, employment and household income were significantly associated with diagnosed diabetes. The current study greatly extends this finding by using nationally representative data, by making comparisons between Indigenous and non‐Indigenous populations, and by using identical SES measures with comparable scales in the two populations. The main limitations of the study relate to the cross‐sectional nature of the study and the potential misclassification of diabetes and SES. Because information on SES and diabetes were collected at the same time, the temporal relationships between SES indicators and diabetes are not always certain. For example, employment status may change as a result of having a serious chronic disease such as diabetes. However, several previous studies have shown a relationship between SES and the incidence of diabetes; in these studies, the temporal relationship between SES and diabetes was clear. Although most participants who said they had diabetes reported that they had been told this by a health practitioner, it is possible that some people who reported diabetes did not actually have it, while others who did have diabetes did not report it. In addition, no information was available about type of diabetes, and it was not possible to distinguish between diabetes and high sugar levels. However, it is unlikely that any such misclassification was systematic across all the various SES measures included, which ranged from individual to household to area‐based measures. Despite being based on self‐report rather than direct measurement, the Australian Institute of Health and Welfare has assessed the National Health Survey data as being the best available indicator of current diagnosed diabetes prevalence in Australia. Information used to determine SES may have been incorrectly reported by some participants, and only limited detail was available on the SES indicators examined here. Given the strong relationship between home ownership and diabetes prevalence in the Indigenous population, it is unfortunate that similar data were not available in the NATSIHS CURF for non‐Indigenous people. Despite the use of comparable scales, the equivalence of a given level of SES may not be guaranteed across individuals or population groups. For example, the meaning of a certain level of education may vary over time and place, and years of education do not necessarily reflect the quality of education received, nor its social or economic value. Similarly, the use of SEIFA quintiles based on the whole population may not adequately capture the socio‐economic position of population subgroups such as Indigenous Australians. No information was available about other potentially important SES measures, such as SES in childhood or total household assets. Although an area‐based measure of disadvantage was included, no other information was available about neighbourhood/area characteristics. Although equivalised household income is intended to adjust for household size and economies of scale, the dynamic nature of Indigenous households can make it difficult to assess both Indigenous household income and household size, both of which are required to calculate equivalised income. Aside from income, the level of missing data was generally small and unlikely to result in substantial bias. Aside from the current study and the previously mentioned DRUID study, few studies have focused on the relationship between SES and diabetes among Indigenous populations in developed countries. One exception is the 1991 Canadian Aboriginal Peoples Survey, in which the prevalence of self‐reported diabetes was inversely associated with income even after adjusting for age. Studies in other minority groups, such as African American women and Black Caribbean and Indian men and Pakistani women in the United Kingdom, have also shown a higher prevalence of diabetes among those of lower SES. There have also been few previous studies of the relationship between SES and other chronic diseases among Indigenous populations. Cass and colleagues showed a strong gradient in regional rates of Indigenous Australian end‐stage renal disease according to an index of social disadvantage. Even in the least disadvantaged regions, the age‐ and sex‐standardised incidence of end‐stage renal disease was generally significantly higher for Indigenous Australians than for the total Australian population. This is consistent with the results of the present study, in which the prevalence of diabetes was substantially higher for Indigenous people than for non‐Indigenous people at every level of SES. Despite tremendous diversity, Australia's Indigenous population is generally treated as a homogeneous group in statistical collections, in large part because limitations in the availability and quality of data prevent further disaggregation. As a consequence, little is known about the distribution of socio‐economic disadvantage within the Indigenous population, or its relationship with health conditions such as diabetes. Recent developments, including the regular implementation of health and social surveys of the Indigenous population, run in parallel by the ABS with the corresponding mainstream collections, have meant that better data are now becoming available. It is imperative that public health researchers and policy makers take advantage of these improvements in data quality and availability, and use the information to gain a more nuanced understanding of their areas of interest and responsibility. For example, Thomas and colleagues recently analysed data on the social determinants of non‐smoking in the Indigenous population and recommended that tobacco control programs consider additional targeting of more disadvantaged groups within the Indigenous population. There is much current interest in actions to ‘Close the Gap’ in Indigenous health and wellbeing. The Council of Australian Governments has committed hundreds of millions of dollars over several years in areas such as housing, economic development and early childhood education. While reducing socio‐economic disadvantage is a legitimate goal in its own right, the data on diabetes suggest that the elimination of existing disparities in SES – a formidable challenge faced by developed nations worldwide – may not necessarily result in a complete elimination of existing health disparities. That is, the results of the current study suggest that even if Indigenous and non‐Indigenous Australians had the same SES, Indigenous Australians would still have a greater prevalence of diabetes, and possibly other chronic diseases as well. If we are to succeed in eliminating health disparities between Indigenous and non‐Indigenous Australians, we must certainly close the socio‐economic gap, but we must also better understand and address the factors that affect Indigenous Australians across the SES spectrum. Acknowledgements I gratefully acknowledge the staff of the Australian Bureau of Statistics for a range of contributions, including the design and implementation of the NATSIHS and NHS, and the development and support of the Remote Area Data Laboratory. I also thank all NATSIHS and NHS participants; this study would not have been possible without them. This work was supported by a National Health and Medical Research Council Research Fellowship (#545200). http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Australian and New Zealand Journal of Public Health Wiley

Socio‐economic gradients in self‐reported diabetes for Indigenous and non‐Indigenous Australians aged 18–64

Download PDF
7 pages

Loading next page...
 
/lp/wiley/socio-economic-gradients-in-self-reported-diabetes-for-indigenous-and-vzGWM5OLj0

References (44)

Publisher
Wiley
Copyright
© 2010 The Authors. Journal Compilation © 2010 Public Health Association of Australia
ISSN
1326-0200
eISSN
1753-6405
DOI
10.1111/j.1753-6405.2010.00547.x
pmid
20618286
Publisher site
See Article on Publisher Site

Abstract

D iabetes is an important cause of morbidity and mortality worldwide. In developed countries, diabetes is more common among those of lower socio‐economic status, but the reverse has been true in some developing countries. The burden of diabetes is particularly pronounced among Indigenous Australians, but little is known about its distribution within the Indigenous population. A recent analysis found significant inverse socio‐economic gradients in diabetes prevalence among urban Indigenous participants in a study in northern Australia, but it is not clear whether this finding is indicative of trends in the Indigenous population elsewhere, or whether any socio‐economic gradients in the Indigenous population are of similar magnitude to those in the non‐Indigenous population. The aim of the current study is to examine socio‐economic gradients in diabetes among a nationally representative sample of Indigenous Australians, and to compare these with corresponding gradients in the non‐Indigenous population. Methods Data for Indigenous and non‐Indigenous adults aged 18–64 years were taken from two national surveys conducted in parallel by the Australian Bureau of Statistics (ABS) in 2004–05: the National Aboriginal and Torres Strait Islander Health Survey (NATSIHS) and the National Health Survey (NHS). These two surveys had very similar content and in most cases the wording of questions on particular topics was identical. This analysis is limited to responses to questions deemed by the ABS to be comparable in the two surveys. The NATSIHS was conducted using two different multi‐stage sampling strategies. In discrete Indigenous communities in remote areas in the Northern Territory, Queensland, South Australia and Western Australia, communities were randomly selected, with probability of selection proportional to community size. A random sample of dwellings within these communities was selected, with one Indigenous adult and up to one Indigenous child then randomly selected to participate. In the remainder of Australia, Census collection districts (CDs) were stratified by state, remoteness classification and the number of Indigenous dwellings in the 2001 Census. A sample of CDs was selected, with probability of selection based on the number of Indigenous households in 2001. Within selected CDs, dwellings were randomly selected and screened to determine whether they included Indigenous members. If they did, then up to two Indigenous adults and up to two Indigenous children were randomly selected to participate. In addition, Indigenous respondents from the NHS were included with NATSIHS data to provide Indigenous population estimates. The NHS was conducted during the same timeframe as the NATSIHS using similar methods and materials. Dwellings were selected using a multi‐stage sampling strategy, and one adult and up to one child were randomly chosen within selected dwellings. Very remote areas were out of scope in the NHS. Both surveys were limited to usual residents of private dwellings, and both were conducted by trained ABS interviewers. In both the NHS and in non‐remote areas in the NATSIHS, data were collected using a Computer Assisted Interview technique. In remote areas of the NATSIHS, pen and paper interview forms were used and some questions were simplified or deleted. Adults aged 18 years and over were personally interviewed. More details about the design and conduct of the surveys have been published elsewhere. To allow data access to interested researchers, the ABS created a Confidentialised Unit Record File (CURF) for the NATSIHS. This file includes unit records for Indigenous respondents of the 2004–05 NATSIHS and the 2004–05 NHS, as well as unit records for non‐Indigenous respondents from the 2004–05 NHS. Although the CURF contains unit records for participants of all ages, this analysis is limited to data from the 20,849 respondents (5,417 Indigenous and 15,432 non‐Indigenous) aged 18–64 years. The exclusion of those aged 65 years and older was due to uncertainty about the applicability of socio‐economic indicators among older people, as well as the relatively small size of this group in the Indigenous population. Definition of diabetes Diabetes was considered to be present if the respondent indicated he or she currently had diabetes or high sugar levels. Diabetes that was reported as not current was not included. All current diabetes was assumed in the surveys to be long‐term (that is, having lasted, or expected to last, at least six months). Almost all respondents who reported diabetes or high sugar levels (referred to hereafter as diabetes) reported that they had been told this by a health care provider. No information was available about type of diabetes, and it was not possible to distinguish between diabetes and high sugar levels as these were not asked about separately in remote areas. Socio‐demographic factors Information was available on a range of socio‐economic and demographic factors, including age, sex, educational attainment, non‐school qualifications, employment status, household income, home ownership (Indigenous respondents only), area of residence and area‐level disadvantage. Available categories for these variables are as shown in Table 1 , although these have been combined for analysis as appropriate. Information about age, sex and whether the respondent was currently attending school was provided by ‘any responsible adult’ within the household; information about the dwelling (including tenure) and the income of non‐participant household members (required to calculate household income) was provided by a household ‘spokesperson’, chosen on the basis of his or her ability to provide accurate information. Information relating to geography (including remoteness classification and area‐level disadvantage score) was provided by the ABS based on the CD in which the selected dwelling was located. All other information used in this analysis was provided by the respondent. 1 Socio‐demographic characteristics of Indigenous and non‐Indigenous adults aged 18–64. a Indigenous % (95% CI) b Non‐Indigenous % (95% CI) b Age (years) 18–24 23.1 (21.7–24.4) 15.1 (14.8–15.4) 25–34 28.4 (27.7–29.0) 22.4 (22.3–22.6) 35–44 24.0 (23.5–24.5) 23.5 (23.4–23.7) 45–54 16.1 (15.7–16.4) 22.0 (21.8–22.1) 55–64 8.5 (7.1–9.9) 17.0 (16.9–17.1) Male 46.8 (45.6–47.9) 49.8 (49.6–50.1) Highest year of school completed Year 12 23.5 (21.2–25.8) 52.5 (51.2–53.8) Year 11 13.0 (11.7–14.4) 10.9 (10.3–11.6) Year 10 31.2 (29.4–33.1) 24.7 (23.7–25.7) Year 9 13.9 (12.5–15.3) 6.3 (5.8–6.7) Year 8 or less 17.3 (15.7–18.9) 5.4 (4.8–5.9) Never went to school 1.1 (0.5–1.6) 0.2 (0.1–0.3) Level of highest non‐school qualification Post‐graduate degree 1.9 (1.0–2.7) 6.2 (5.8–6.7) Bachelor's degree 2.9 (2.3–3.6) 14.5 (13.8–15.3) Diploma 4.7 (3.7–5.7) 9.7 (9.1–10.3) Certificate 24.2 (22.2–26.1) 26.0 (25.0–27.1) No qualifications 66.4 (64.1–68.6) 43.5 (42.5–44.5) Employment status Employed full‐time 32.7 (30.4–35.1) 54.3 (53.4–55.2) Employed part‐time 21.9 (20.1–23.8) 21.8 (21.1–22.5) Unemployed 8.1 (6.9–9.2) 3.0 (2.7–3.4) Not in the labour force 37.3 (35.0–39.6) 20.9 (20.1–21.7) Housing tenure Owner‐occupied without mortgage 6.8 (5.4–8.3) n/a c Owner‐occupied with mortgage 17.9 (15.7–20.0) n/a c Renter or other tenure 75.3 (72.7–77.9) n/a c Equivalised household income quintile d 1 (lowest) 33.7 (31.4–36.1) 11.3 (10.7–11.9) 2 21.6 (19.7–23.6) 13.1 (12.5–13.8) 3 14.3 (12.4–16.1) 16.9 (16.1–17.6) 4 9.4 (7.7–11.2) 19.5 (18.7–20.2) 5 (highest) 5.2 (4.0–6.4) 21.7 (20.7–22.7) Not known or not stated 15.6 (13.6–17.6) 17.5 (16.6–18.4) SEIFA quintile e 1 (most disadvantaged) 49.3 (43.7–55.0) 17.1 (15.7–18.5) 2 19.3 (15.2–23.3) 19.0 (17.4–20.7) 3 18.5 (14.3–22.7) 20.3 (18.4–22.2) 4 9.0 (6.4–11.6) 21.3 (19.5–23.0) 5 (least disadvantaged) 3.9 (2.2–5.7) 22.3 (20.0–24.7) Area of residence Major cities 30.6 (29.1–32.0) 70.2 (68.6–71.8) Inner regional 20.1 (19.0–21.3) 19.5 (17.9–21.0) Outer regional 21.5 (20.4–22.5) 10.4 (9.2–11.5) Remote or very remote 27.8 (26.3–29.4) — Notes: a) Source: Weighted data from the National Aboriginal and Torres Strait Islander Health Survey 2004–05 confidentialised unit record file (CURF). b) CI, confidence interval. Proportions are weighted to provide population estimates. Totals are based on those with non‐missing data, except for equivalised household income, for which a separate category is shown. c) n/a, not available. d) Gross weekly equivalised cash income of household, using the OECD scale. Quintiles are based on national figures. e) SEIFA, Socioeconomic Index for Areas, Index of Relative Disadvantage. Those reported to be still at school (n=67) were not asked questions about the highest year of school completed or non‐school qualifications. They have been coded as missing on both of these variables. Those whose educational attainment was not stated (n=2) and those whose level of qualifications could not be determined (n=222) were coded as missing on the relevant variable only. Gross weekly household equivalised income, which takes into account household size and composition, was estimated using the OECD scale. Quintiles were determined based on all‐Australian figures. That is, the same categories were used for both Indigenous and non‐Indigenous participants. Equivalised income quintile was not available for 2,941 respondents (14.1%). Analyses were conducted with these respondents coded as missing, as well as with them included using a special category of household income unknown. Home ownership was available in the CURF only for Indigenous respondents (missing for n=41), and was based on whether the home was owned or being purchased by any of its occupants (not necessarily the respondent). Area of residence was classified according to the Australian Standard Geographical Classification remoteness classification (based on the ARIA+ index) into major cities, inner regional, outer regional, remote and very remote. ABS documentation indicates that the remote/very remote category was to be used for Indigenous respondents only. Therefore, area of residence was re‐coded to missing for 312 non‐Indigenous respondents (2%) whose residence was listed as remote/very remote. Area‐level disadvantage was based on the 2001 socio‐economic Indexes for Areas (SEIFA) Index of Disadvantage score for the CD of the selected dwelling. Quintiles were determined based on all‐Australian figures. That is, the same categories were used for both Indigenous and non‐Indigenous respondents. Those with SEIFA quintile not known (n=313) were coded as missing. Statistical analysis All analyses were conducted using STATA version 10.0 via the ABS's Remote Access Data Laboratory (RADL). Under the RADL system, analysts submit statistical code to the ABS; the code is then run and the output made available to the analyst through a password‐protected web account. Analysts do not have direct access to unit record data at any time, and there are limits placed on the commands and outputs that are allowed, in order to protect the security and confidentiality of the data. All analyses used ABS‐generated person‐weights (or expansion factors) to adjust for disproportionate sampling of some groups. The estimates produced in this manner apply to the population as a whole, and not just the sample. Standard errors and 95% confidence intervals (CI) were calculated using replicate weights produced by the ABS using the Jackknife method (250 replicate weights for Indigenous respondents, 60 for non‐Indigenous respondents). These replicate weights allow the estimation of standard errors taking into account the complex design and weighting procedures used in the surveys. Although STATA version 10 incorporates a suite of procedures to analyse complex survey data, these commands are not allowed in the RADL system (Therese Lalor, ABS, personal communication, May 2009). Instead, commands from the svr module written by Nick Winter (available using the STATA command: search svr, net) were used. STATA code is available from the author on request. Directly age‐standardised estimates and 95% CIs were calculated using an alternative set of person‐weights and replicate weights produced by the ABS for that purpose. The standard population was the total Australian population as at 30 June 2001. Logistic regression was conducted separately for Indigenous and non‐Indigenous groups due to the different numbers of replicate weights for the two groups. All models were adjusted for age group and sex, with socio‐economic variables assessed individually and in combination. Household income and SEIFA score were both modeled using all five quintiles as well as comparing quintiles one and two with quintiles 3–5 combined, but only the latter results are presented. There were relatively few Indigenous participants in the top quintiles of these variables ( Table 1 ), and preliminary analysis indicated a similar prevalence of diabetes in the top three quintiles on both measures, allowing them to be combined with minimal loss of information. Participants with missing data were excluded only from analyses involving the variable for which they were missing data. Ethics approval This study was approved by the Human Research Ethics Committee of the Northern Territory Department of Health and Families and the Menzies School of Health Research. The proposal was considered and approved by both the Aboriginal sub‐committee and the main committee. Results Almost one in 10 Indigenous people (9.7%, 95% CI 8.6%‐10.8%) aged 18–64 reported they currently had diabetes or high blood sugar levels. This was much higher than for the non‐Indigenous population (3.0%, 95% CI 2.6–3.3), even before adjusting for the younger age distribution of the Indigenous population. Diabetes increased more sharply with age in the Indigenous population than in the non‐Indigenous population ( Figure 1 ). There was a 30‐fold increase in prevalence from the youngest to the oldest age group among Indigenous people, about twice that seen in the non‐Indigenous population. 1 Prevalence of self‐reported diabetes by age and Indigenous status, 2004/05. The socio‐demographic profile of the Indigenous population was significantly different from that of the non‐Indigenous population ( Table 1 ), with a younger age distribution, lower educational attainment, and greater levels of disadvantage across a range of indicators. After adjusting for age, diabetes prevalence was significantly higher among those of lower SES in both the Indigenous and non‐Indigenous populations ( Figures 2a–2g , Table 2 ). Area of residence was significantly associated with diabetes prevalence in the Indigenous population, but not in the non‐Indigenous population. The relative differences between low and high SES groups, as measured by age‐ and sex‐adjusted odds ratios, were generally of similar magnitude for the Indigenous and non‐Indigenous populations ( Table 2 ). For example, the relative odds of diabetes for the lowest quintile of equivalised household income (compared with quintiles 3–5 combined) was 2.3 (95% CI 1.6–3.4) for the Indigenous population and 2.0 (95% CI 1.5–2.8) for the non‐Indigenous population. The relative odds of diabetes for those who did not complete at least Year 10 (compared with those who did) was 1.8 for both the Indigenous (95% CI 1.4–2.3) and non‐Indigenous (95% CI 1.4–2.4) populations. Despite these similarities in relative terms, there were marked disparities in absolute terms, with Indigenous people of high SES having a greater prevalence of diabetes than non‐Indigenous people of low SES on every SES measure examined ( Figures 2a–2g ). 2a–c Age‐standardised prevalence of self‐reported diabetes by indicators of socio‐economic status, Indigenous and non‐Indigenous adults aged 18–64, 2004/05. 2d–g Age‐standardised prevalence of self‐reported diabetes by indicators of socio‐economic status, Indigenous and non‐Indigenous adults aged 18–64, 2004/05. 2 Relative odds of diabetes by socioeconomic status for Indigenous and non‐Indigenous adults aged 18–64. a Indigenous Adjusted OR (95% CI) b Non‐Indigenous Adjusted OR (95% CI) b Highest year of school completed Year 10 or more 1.0 1.0 Less than Year 10 c 1.8 (1.4–2.3) 1.8 (1.4–2.4) Level of highest non‐school qualification Bachelor's or post‐graduate degree 0.4 (0.2–0.9) 0.8 (0.6–1.1) Diploma 0.5 (0.2–1.3) 0.5 (0.3–0.8) Certificate 0.6 (0.4–0.8) 0.9 (0.7–1.2) No qualifications 1.0 1.0 Employment status Employed 1.0 1.0 Unemployed 1.9 (1.0–3.4) 0.9 (0.3–2.1) Not in the labour force 1.6 (1.2–2.2) 1.9 (1.4–2.6) Housing tenure Owner‐purchaser 0.4 (0.2–0.5) — Renter or other tenure 1.0 — Equivalised household income quintile d 1 (lowest) 2.3 (1.6–3.4) 2.0 (1.5–2.8) 2 1.7 (1.0–2.7) 1.5 (1.1–2.1) 3–5 (highest) 1.0 1.0 Not known or not stated 2.2 (1.3–3.6) 1.0 (0.7–1.4) SEIFA quintile e 1 (most disadvantaged) 1.6 (1.0–2.5) 2.1 (1.6–2.9) 2 0.9 (0.6–1.5) 1.5 (1.0–2.1) 3–5 (least disadvantaged) 1.0 1.0 Area of residence Major cities 1.0 1.0 Inner regional 1.2 (0.7–1.9) 1.1 (0.8–1.4) Outer regional 1.6 (1.0–2.3) 1.0 (0.8–1.3) Remote or very remote 2.5 (1.8–3.6) — Notes: a) Source: Weighted data from the National Aboriginal and Torres Strait Islander Health Survey 2004–05 confidentialised unit record file (CURF). b) OR, odds ratio. CI, confidence interval. All odds ratios are from logistic regression models adjusted for age group, sex and the individual variable listed. c) Includes those who never went to school. d) Gross weekly equivalised cash income of household, using the OECD scale. Quintiles are based on national figures. e) SEIFA, Socioeconomic Index for Areas, Index of Relative Disadvantage. Adjusting for multiple SES variables simultaneously resulted in some attenuation of the odds ratios, but the directions of the associations remained the same (data not shown). Among Indigenous people, home ownership retained a strong and significant inverse association with diabetes, even after adjustment for other SES variables (OR 0.4, 95% CI 0.3–0.7). Discussion The results presented here indicate that socio‐economic disadvantage explains some but not all of the higher prevalence of diabetes among Indigenous Australians. There are marked SES‐related gradients in self‐reported diabetes among both Indigenous and non‐Indigenous Australians, and Indigenous Australians are over‐represented in lower SES categories. However, the prevalence of diabetes is higher for Indigenous people than non‐Indigenous people in every SES group, often by a considerable margin. Indeed, for most variables considered, Indigenous Australians of high SES are more likely to report diabetes than non‐Indigenous Australians of low SES. The combination of a socio‐economic gradient within the Indigenous population and a gap between the Indigenous and non‐Indigenous populations suggests that traditional risk factors may not be sufficient to explain completely the patterns of diabetes among Indigenous Australians. Other factors that may operate across the socio‐economic spectrum, including racism and discrimination, stress, and a legacy of grief, loss and dispossession, may also play a role through a range of neuroendocrine, autonomic, metabolic, immune and/or behavioural pathways. A recent study indicated that racial disparities in the US may be explained by differences in the ‘health risk’ environments in which African Americans and whites live, and this could be relevant in Australia. Differential access to services may also exist across SES groups. Although genes clearly play a role in the development of diabetes, the relationship between ethnicity and genetic susceptibility is quite complex. These data confirm the relationships previously observed in the DRUID Study among Indigenous people in one northern Australian city, in which home ownership, employment and household income were significantly associated with diagnosed diabetes. The current study greatly extends this finding by using nationally representative data, by making comparisons between Indigenous and non‐Indigenous populations, and by using identical SES measures with comparable scales in the two populations. The main limitations of the study relate to the cross‐sectional nature of the study and the potential misclassification of diabetes and SES. Because information on SES and diabetes were collected at the same time, the temporal relationships between SES indicators and diabetes are not always certain. For example, employment status may change as a result of having a serious chronic disease such as diabetes. However, several previous studies have shown a relationship between SES and the incidence of diabetes; in these studies, the temporal relationship between SES and diabetes was clear. Although most participants who said they had diabetes reported that they had been told this by a health practitioner, it is possible that some people who reported diabetes did not actually have it, while others who did have diabetes did not report it. In addition, no information was available about type of diabetes, and it was not possible to distinguish between diabetes and high sugar levels. However, it is unlikely that any such misclassification was systematic across all the various SES measures included, which ranged from individual to household to area‐based measures. Despite being based on self‐report rather than direct measurement, the Australian Institute of Health and Welfare has assessed the National Health Survey data as being the best available indicator of current diagnosed diabetes prevalence in Australia. Information used to determine SES may have been incorrectly reported by some participants, and only limited detail was available on the SES indicators examined here. Given the strong relationship between home ownership and diabetes prevalence in the Indigenous population, it is unfortunate that similar data were not available in the NATSIHS CURF for non‐Indigenous people. Despite the use of comparable scales, the equivalence of a given level of SES may not be guaranteed across individuals or population groups. For example, the meaning of a certain level of education may vary over time and place, and years of education do not necessarily reflect the quality of education received, nor its social or economic value. Similarly, the use of SEIFA quintiles based on the whole population may not adequately capture the socio‐economic position of population subgroups such as Indigenous Australians. No information was available about other potentially important SES measures, such as SES in childhood or total household assets. Although an area‐based measure of disadvantage was included, no other information was available about neighbourhood/area characteristics. Although equivalised household income is intended to adjust for household size and economies of scale, the dynamic nature of Indigenous households can make it difficult to assess both Indigenous household income and household size, both of which are required to calculate equivalised income. Aside from income, the level of missing data was generally small and unlikely to result in substantial bias. Aside from the current study and the previously mentioned DRUID study, few studies have focused on the relationship between SES and diabetes among Indigenous populations in developed countries. One exception is the 1991 Canadian Aboriginal Peoples Survey, in which the prevalence of self‐reported diabetes was inversely associated with income even after adjusting for age. Studies in other minority groups, such as African American women and Black Caribbean and Indian men and Pakistani women in the United Kingdom, have also shown a higher prevalence of diabetes among those of lower SES. There have also been few previous studies of the relationship between SES and other chronic diseases among Indigenous populations. Cass and colleagues showed a strong gradient in regional rates of Indigenous Australian end‐stage renal disease according to an index of social disadvantage. Even in the least disadvantaged regions, the age‐ and sex‐standardised incidence of end‐stage renal disease was generally significantly higher for Indigenous Australians than for the total Australian population. This is consistent with the results of the present study, in which the prevalence of diabetes was substantially higher for Indigenous people than for non‐Indigenous people at every level of SES. Despite tremendous diversity, Australia's Indigenous population is generally treated as a homogeneous group in statistical collections, in large part because limitations in the availability and quality of data prevent further disaggregation. As a consequence, little is known about the distribution of socio‐economic disadvantage within the Indigenous population, or its relationship with health conditions such as diabetes. Recent developments, including the regular implementation of health and social surveys of the Indigenous population, run in parallel by the ABS with the corresponding mainstream collections, have meant that better data are now becoming available. It is imperative that public health researchers and policy makers take advantage of these improvements in data quality and availability, and use the information to gain a more nuanced understanding of their areas of interest and responsibility. For example, Thomas and colleagues recently analysed data on the social determinants of non‐smoking in the Indigenous population and recommended that tobacco control programs consider additional targeting of more disadvantaged groups within the Indigenous population. There is much current interest in actions to ‘Close the Gap’ in Indigenous health and wellbeing. The Council of Australian Governments has committed hundreds of millions of dollars over several years in areas such as housing, economic development and early childhood education. While reducing socio‐economic disadvantage is a legitimate goal in its own right, the data on diabetes suggest that the elimination of existing disparities in SES – a formidable challenge faced by developed nations worldwide – may not necessarily result in a complete elimination of existing health disparities. That is, the results of the current study suggest that even if Indigenous and non‐Indigenous Australians had the same SES, Indigenous Australians would still have a greater prevalence of diabetes, and possibly other chronic diseases as well. If we are to succeed in eliminating health disparities between Indigenous and non‐Indigenous Australians, we must certainly close the socio‐economic gap, but we must also better understand and address the factors that affect Indigenous Australians across the SES spectrum. Acknowledgements I gratefully acknowledge the staff of the Australian Bureau of Statistics for a range of contributions, including the design and implementation of the NATSIHS and NHS, and the development and support of the Remote Area Data Laboratory. I also thank all NATSIHS and NHS participants; this study would not have been possible without them. This work was supported by a National Health and Medical Research Council Research Fellowship (#545200).

Journal

Australian and New Zealand Journal of Public Health – Wiley

Published: Jul 1, 2010

There are no references for this article.