Title Page
Article Title:
Trends in chronic disease mortality in the Northern Territory Aboriginal population, 19972004: using underlying and multiple causes of death.
WORD COUNT
Abstract = 254
Main text (including references) = 2783
1
Trends in chronic disease mortality in the Northern Territory Aboriginal population,
1997-2004: using underlying and multiple causes of death
Emily Fearnley, Shu Qin Li, Steven Guthridge
Abstract
Objective: To assess trends in chronic disease mortality in the Aboriginal population of the
Northern Territory (NT), using both underlying and multiple causes of death.
Method: Death registration data from 1997 to 2004, were used for the analysis of deaths from
five chronic diseases; ischaemic heart disease (IHD), diabetes, chronic obstructive pulmonary
disease (COPD), renal failure and stroke. Negative binomial regression models were used to
estimate the average annual change in mortality rates for each of the five diseases. Chi
squared tests were conducted to determine associations between the five diseases.
Results: The five chronic diseases contributed to 49.3% of all Aboriginal deaths in the NT.
The mortality rate ratio of NT Aboriginal to all Australian death rates from each of the
diseases ranged from 4.3 to 13.0, with the lowest rate ratio for stroke and highest for diabetes.
There were significant statistical associations between IHD, diabetes, renal failure and stroke.
The mortality rates for diabetes, COPD and stroke declined at estimated annual rates for NT
Aboriginal males of 3.6%, 1.0% and 11.7% and for Aboriginal females by 3.5%, 6.1% and
7.1% respectively. There were increases in mortality rates for Aboriginal males and females
for IHD and a mixed result for renal failure.
Conclusion: NT Aboriginal people experience high chronic disease mortality; however
mortality rates appear to be declining for diabetes, COPD and stroke. The impact of chronic
2
disease on mortality is greater than previously reported by using a single underlying cause of
death. The results highlight the importance of integrated chronic disease interventions.
Introduction
An epidemic of chronic disease has developed in Aboriginal Australians commencing in the
early 1980s (1). The mortality rate in the Northern Territory (NT) Aboriginal population has
been reported as greater than for other Australians (2-4), and although declines have been
observed in the overall rate of NT Aboriginal mortality over four decades (1966 – 2001), the
reduction has been less than in the total Australian population (2). A recent analysis of longterm trends in NT Aboriginal chronic disease mortality identified a significant increase in
mortality rates for ischaemic heart disease (IHD) and diabetes over the 25 years, from 1977 to
2001 (5). Another study assessed the cause of the life expectancy gap between NT Aboriginal
and non-Aboriginal people, and highlighted that through the 80’s and 90’s there was an
increasing contribution from non-communicable and lifestyle diseases, which has offset a
declining contribution from communicable diseases (4). Other authors have noted that as life
expectancy increases, the major health conditions in a population shift towards chronic
diseases (6). A characteristic of chronic diseases is that many have shared risk factors and
frequently co-exist with other chronic diseases (6,7). Assessing the co-occurrence of chronic
diseases helps to identify those diseases that are more likely to occur together, allowing the
design of appropriate and targeted interventions.
Historically, mortality studies have relied on the use of a single, underlying cause of death;
which is the disease or condition that led directly to death (8). This provides useful, but
limited information particularly for chronic diseases such as diabetes which are less likely to
3
be reported as the underlying cause of death (8). As a result, the contribution of chronic
diseases to total mortality has been under-estimated. In 1997, the Australian Bureau of
Statistics (ABS) introduced coding for multiple or associated causes of death in death
registration data (8). The use of multiple causes of death coding in mortality data provides the
opportunity to gain a more comprehensive view of health within a population, including
information about the interaction of diseases and of the contributing causes of death (7,
9,10,11). Reports which utilise multiple causes of death are starting to appear in the literature,
including a recent analysis of all-cause and cause-specific death rates for the NT Aboriginal
population by varying levels of geographic remoteness (12).
In recent years, concerted efforts have been made in the NT to tackle chronic diseases, with
strategies including early screening and active clinical intervention (13). A systematic chronic
disease strategy was established in the NT in 1999 and it is timely to assess changes that may
have occurred in association with the more focused and collective efforts. This study
examines the contribution of five common chronic diseases within the NT by using both
underlying and multiple causes of death. The study also analyses time trends in morality rates
and the associations between each of the five selected chronic diseases as causes of death.
Methods
Data for this study were obtained from two sources: NT and Australian death data from the
ABS death registration data, and NT and Australian estimated resident population from the
ABS 2001 Census of Population and Housing. The NT deaths included in this study were the
deaths of all NT residents occurring in Australia, with separate analysis for the Aboriginal and
non-Aboriginal population. The quality of Indigenous status in death registration and
population data within the NT is high and has been previously reported as acceptable for trend
4
analysis (14). From 1997, the ABS has classified both underlying and multiple causes of
death, by using the International Classification of Diseases (ICD 10) tenth revision for death
registration (8). The five common, chronic diseases included in this study are ischaemic heart,
disease (IHD), diabetes, stroke, chronic obstructive pulmonary disease (COPD) and renal
failure. The ICD 10 codes used are listed in Table 1.
Total numbers and percentages of deaths were calculated for all five chronic diseases, by
underlying and multiple causes of death. NT mortality rates for each disease were calculated
using direct standardisation and age-adjusted to the Australian standard population (2001).
Chi squared analyses were conducted using two by two tables to determine associations
between chronic diseases. The Bonferroni method was used to adjust the cut off for statistical
significance testing for multiple hypothesis tests. Age-adjusted mortality rate ratios (MRR)
were used to determine differences in rates in the NT Aboriginal population compared to the
Australian population. A negative binomial regression model was used to estimate the
average annual change in mortality rate for each of the five chronic diseases for the NT
Aboriginal population compared with NT non-Aboriginal population and the Australian
population. Various interaction terms were added to the negative binomial regression model
to test differences in annual changes between the NT Aboriginal population and all
Australians . All statistical analysis was conducted using Stata software (version 9.0;
StataCorp, College Station, Texas, USA).
The project was approved by the Human Research Ethics Committee of the Department of
Health and Community Services and Menzies School of Health Research (Project number
07/69), and the Australian National University Human Research Ethics Committee (Protocol
number 2007/2265).
5
Results
Between 1997 and 2004, there were a total of 3438 deaths among NT Aboriginal residents. Of
this total, 1115 (32.4%) deaths listed one of the five selected chronic diseases assessed in this
paper as the underlying cause of death. A further 579 (16.8%) deaths listed at least one of the
five chronic diseases as a multiple cause of death, without another of the five diseases listed
as the underlying cause of death. A total of 1694 (49.3%) NT Aboriginal deaths had at lest
one of the five chronic diseases listed as either the underlying or multiple cause of death.
Using both underlying and multiple causes, IHD was identified as the leading chronic disease
that contributed to deaths, with 711 (20.7%) deaths followed by diabetes (524, 15.2%), renal
failure (518, 15.1%), COPD (396, 11.5%), and stroke (299, 8.7%).
Among Aboriginal people who died of IHD, 474 (66.7%) had IHD recorded as the underlying
cause of death and for another 237 (33.3%), IHD was recorded as one of the multiple causes
of death. In contrast, diabetes, renal failure, COPD and stroke were more likely to be recorded
as a multiple cause than underlying cause of death. The ratio of multiple to underlying cause
of death for diabetes, renal failure, COPD and stroke were 1.3, 4.6, 1.2 and 1.1 respectively
(Table 2). Deaths due to renal disease were the least common of the five chronic diseases to
be reported as the underlying cause of death, with only 93 deaths listing renal disease as the
underlying cause, but an additional 425 deaths listed renal disease as a multiple cause of
death.
Figures 1 and 2 outline the age-adjusted mortality rates for five chronic diseases by all
contributing causes of death, for NT Aboriginal population compared with NT nonAboriginal population and all Australian. The age-adjusted mortality rates for all five chronic
6
diseases were much higher in the NT Aboriginal than both NT non-Aboriginal and all
Australian populations, especially for diabetes and renal disease in Aboriginal females.
The co-occurrence of the five chronic diseases by Indigenous status is presented in Table 3. A
significantly high level (P<0.0127) of association was observed between a number of
diseases, including IHD and diabetes, IHD and renal failure, IHD and stroke, diabetes and
stroke, diabetes and renal failure.
The age adjusted mortality rate ratio (MRR) of NT Aboriginal deaths to all Australian deaths
was statistically significant for all five chronic diseases, using all contributing causes of death
and ranged from the lowest 4.3(3.5, 5.1) for stroke to the highest 13.0, (11.3, 14.9) for
diabetes (Table 4). Similarly, the mortality rate ratios of NT Aboriginal deaths to NT nonAboriginal deaths were also significantly higher for all five chronic diseases, and ranged from
4.1 (3.3, 5.0) for stroke to 10.3 (8.3, 12.8) for renal failure
The negative binomial regression model estimated average annual changes in all contributing
causes of death rate for each of the five chronic diseases are presented in Table 5. While most
changes are not statistically significant they indicate important and sometimes substantial
changes. The mortality rate for IHD increased in the NT Aboriginal population, with an
annual increase of 1.3% (-5.0, 8.1) in Aboriginal males and 2.9 %(-5.2, 11.7) for Aboriginal
females compared to reductions of 3.5%,(-7.2, 0.2) in Australian males and 3.2% (-7.6, 1.4) in
females. There was a greater reduction in the annual mortality rate for diabetes in the NT
Aboriginal males (-3.6%, (-8.8, 1.9)) and NT Aboriginal females (-3.5%,(-9.4, 2.9)) compared
with the smaller declines for Australian males (-0.6%, (-1.6,0.3)) and Australian females (1.3%, (-4.2, 1.7) .The mortality rate from COPD declined for both the NT Aboriginal
7
population and all Australians, with average annual declines of 1.0% (-9.6, 8.4) in NT
Aboriginal males and 6.1 (-12.3, 0.6) compared with a decline of 5.0% (-10.2, 0.5) for
Australian males and 2.8% (-4.5, 1.0) for Australian females. The mortality rate from stroke
declined in both the NT Aboriginal population and all Australians population, with a greater
reduction in average annual mortality rate 11.7%, (-19.5, -3.1) for NT Aboriginal males and
7.1% (-15.8, 2.4) for NT Aboriginal females. There was a mixed pattern observed for death
from renal failure in the NT Aboriginal population. The estimated annual reduction rate in
NT Aboriginal males was 1.3% (-9.4, 7.6) whereas there was an annual average increase of
4.4% (-3.7, 13.2) in NT Aboriginal females. These rates compare with the annual reductions
of 4.5% (-9.4, 7.6) in Australian males and 2.0% (-7.2, 3.5) in Australian females.
Discussion
The combination of underlying and multiple causes of death provides a more comprehensive
overview of mortality, including a more complete assessment of the burden of chronic
diseases within a population, than available from the sole use of underlying cause of death
data (7,8,9,10,11). The previous underestimation of the impact of a specific disease is
especially evident within this study for renal disease, for which the burden of the disease was
more than four times greater if multiple causes of death were considered compared to limiting
analysis to the underlying cause of death. Previous studies have highlighted that certain
chronic diseases; particularly diabetes, COPD, stroke, and renal disease are frequently
overlooked as the underlying cause of death, a shortcoming that is addressed by use of
multiple cause of death data (7, 9). This study highlights the substantial contribution these
chronic diseases made to the overall mortality in the NT Aboriginal population, and the
associated need for appropriate health policy and action to target the prevention and reduction
of chronic diseases.
8
The poor health status and higher mortality rate of the NT Aboriginal population has been
well documented (2-4). There are many factors that contribute to the higher chronic disease
mortality rates including social disadvantage, poor access to primary health care and high
prevalence of smoking (6, 15). This study emphasises the association between chronic
diseases, with people who had one of the selected chronic diseases as a cause of death more
likely to have had other chronic diseases. This is partially explained by the shared risk factors
of chronic diseases including poor nutrition, lack of exercise, smoking and high alcohol intake
(6). Most of the selected chronic diseases were significantly associated with one another.
The co-occurrence of chronic diseases reinforces the importance of an integrated approach to
tackle common chronic diseases. The Northern Territory Department of Health and Families
developed the first integrated non-communicable diseases strategy in Australia (13) and
collective efforts have been made for more than a decade to reduce the morbidity and
mortality of chronic diseases in the Northern Territory through screening, prevention and
clinical management.
A recent study shows improvements in life expectancy in the NT Aboriginal population in the
recent decades (16). Another NT study reports that between 1990 and 2001, the death rate
from some chronic diseases such as COPD and stroke declined in the NT Aboriginal
population, but increased for IHD and diabetes (5). Using both underlying and multiple cause
of death, and more recent data, the current study provides a more comprehensive picture of
mortality ascribed to chronic disease in the NT Aboriginal population. This study found that
the death rate from diabetes, COPD and stroke have all declined between 1997 and 2004 in
the NT Aboriginal population and a decline in mortality rate was also observed in renal
failure death in NT Aboriginal males. Although not statistically significant, the reduction in
9
annual mortality rate in diabetes and stroke were much higher in the NT Aboriginal
population and this may indicate early signs of improvement in chronic disease mortality. The
results support an argument that collective efforts in early screening and treatment of chronic
diseases are resulting in improved health outcomes in the NT. There remain areas of
continuing concern, particularly the increasing death rates from IHD in NT Aboriginal males
and females and from renal failure in NT Aboriginal females. A recent paper demonstrated
the complexity of the challenge to reduce mortality rates, when reporting that improved
survival after Acute Myocardial Infarction (AMI) for the NT Aboriginal population was
counterbalanced by the increasing AMI incidence (17). It is likely that parallel increases in
other chronic diseases are placing a similar brake on improved outcomes.
The limitations of this study include the reliance on data collection from death certificates,
which have a number of imperfections, including that coding practices are known to change
over time (9, 18). Another limitation of this study is the relatively short time period that data
on multiple cause of death has been available. This period was constrained on one hand by the
introduction of multiple causes of death coding in 1997, and on the other by the availability of
data up until 2004, at the time of the study. The death registration and coding process is often
lengthy and incomplete until a number of years after deaths have occurred.
References
1. Australian Bureau of Statistics and Australian Institute of Health and Welfare. The
Health and welfare of Australia’s Aboriginal and Torres Strait Islander peoples 2005.
Canberra (ACT): ABS; 2005. Catalogue. No.: 4704.0.
2. Condon JR, Barnes T, Cunningham J, Smith L. Improvements in Indigenous mortality
in the Northern Territory over four decades. Aust N Z J Public Health 2004;28:445-51.
10
3. Cunningham J, Paradies Y. Mortality of Aboriginal and Torres Strait Islander
Australians 1997. Canberra (ACT): Australian Bureau of Statistics; 2000. Occasional
Paper No.: 3315.0.
4. Zhao Y, Dempsey K. Causes of inequality in life expectancy between Indigenous and
non-Indigenous people in the Northern Territory, 1981-2000: a decomposition
analysis. Med J Aust 2006;184:490-4.
5. Thomas DP, Condon JR, Anderson IP, Li SQ, Halpin S, Cunningham J, et al. Longterm trends in Indigenous deaths from chronic diseases in the Northern Territory: a
foot on the brake, a foot on the accelerator. Med J Aust 2006;185:145-9.
6. Condon JR, Warman G, Arnold L. The health and welfare of Territorians. Darwin
(NT): Territory Health Services; 2001. p. 97-104.
7. Tardon AG, Zaplana J, Hernandez R, Cueto A. Usefulness of the codification of
multiple causes of death in mortality statistics. Int J Epidemiol 1995;24:1132-7.
8. Australian Bureau of Statistics. Multiple cause of death analysis, 1997-2001.
Canberra (ACT): ABS; 2003. Catalogue. No.: 3319.0.55.001.
9. Li SQ, Cunningham J, Cass A. Renal-related deaths in Australia 1997-1999. Intern
Med J 2004;34:259-65.
10. Redelings MD, Sorvillo F, Simon P. A comparison of underlying cause and multiple
causes of death. Epidemiology 2006;17:100-3.
11. Mackenbach JP, Kunst AE, Lautenbach H, Bijlsma F, Oei YB. Competing causes of
death: An analysis using multiple cause of death data from The Netherlands. Am J
Epidemiol 1995;141:466-75.
12. Andreasyan K, Hoy WE. Patterns of mortality in Indigenous adults in the Northern
Territory, 1998-2003: are people living in more remote areas worse off? Med J Aust
2009;190:307-11.
11
13. Weeramanthri T, Morton S, Hendy S, Connors C, Rae C, Ashbridge D. Northern
Territory preventable chronic disease strategy – Overview and framework. Darwin
(NT): Territory Health Services; 1999.
14. Condon JR, Barnes T, Cunningham J, Smith L. Demographic characteristics and
trends of Northern Territory Indigenous population, 1966 to 2001. Darwin (NT):
Cooperative Research Centre for Aboriginal Health. Occasional Paper. 2004.
15. Cunningham J. Cigarette smoking among Indigenous Australians, 1994. Canberra
(ACT): Australian Bureau of Statistics; 1997. Occasional Paper Catalogue No.:
4701.0.
16. Wilson T, Condon JR, Barnes T. Northern Territory Indigenous life expectancy
improvements, 1967-2004. Aust N Z J Public Health 2007;31:184-8.
17. You JQ, Condon J, Zhao Y, Guthridge S. Incidence and survival analysis for acute
myocardial infarctions in the Northern Territory, 1992–2004. Med J Aust .MJA
2009; 190 (6): 298-302
18. Mckenzie K, Tong S, Walker S, Sadkowsky K. Evolution in classifying mortality
statistics. Canberra. The National Centre for Classification in Health.2002.
12
Table 1 ICD 10 codes for IHD, diabetes, stroke, COPD and renal failure
Conditions
Ischaemic heart disease
Diabetes
Stroke
COPD
Renal failure
ICD 10 Codes
I20, I21, I22, I23, I24, I25
E10, E11, E13, E14
I6
J40, J41, J42, J43, J44, J47
N17, N18, N19
Table 2 Number of underlying and multiple causes of death reported for selected five chronic
diseases by Indigenous status, NT population, 1997 – 2004.
Underlying cause of death
IHD
Diabetes
COPD
Renal
Stroke
NT
Aboriginal
474
226
180
93
142
NT nonAboriginal Total
525 999
86 312
173 353
25 118
175 317
Multiple cause of death
NT
Aboriginal
237
298
216
425
157
NT nonAboriginal Total
279 516
190 488
220 436
219 644
144 301
Ratio of Multiple to Underlying
NT
Aboriginal
0.5
1.3
1.2
4.6
1.1
NT nonAboriginal Total
0.5 0.51
2.2 1.56
1.3 1.23
8.8
5.5
0.8
0.9
13
Figure 1 Age-adjusted mortality rates per 100 000 population for selected chronic diseases,
males, 1997-2004.
NT Aboriginal males
NT non-Aboriginal males
COPD
per 100 000 population
Diabetes
Age adjusted death rate
Australia males
600
500
400
300
200
Stroke
Renal
0
IHD
100
Figure 2 Age-adjusted mortality rates per 100 000 population for selected chronic diseases,
females, 1997-2004.
Age adjusted death
rate, per 100 000
population
NT Aboriginal females
NT non-Aboriginal females
Australia females
600
500
400
300
200
100
Stroke
Renal
COPD
Diabetes
IHD
0
14
Table 3 The association of five chronic diseases as reported causes of death, by Indigenous
status, Northern Territory 1997– 2004
Co-conditions for IHD deaths
Diabetes
Renal
COPD
Stroke
Aboriginal % (p)
27.6 (0.000)*
18.6 (0.003)*
13.7 (0.045)
5.5 (0.001)*
Non-Aboriginal % (p)
17.9 (0.000)*
10.7 (0.000)*
15.6 (0.000)*
7.5 (0.042)
Co-conditions for diabetes deaths
IHD
Renal
Stroke
COPD
Co-conditions for COPD deaths
IHD
Renal
Diabetes
Stroke
Aboriginal % (p)
24.5 (0.045)
15.2 (0.964)
12.9 (0.163)
6.8 (0.158)
Non-Aboriginal % (p)
31.8 (0.000)*
Diabetes
6.6 (0.688)
IHD
11.2 (0.014)
COPD
7.9 (0.311)
Stroke
Aboriginal % (P) Non-Aboriginal % (p)
37.4 (0.000)*
52.2 (0.000)*
34.5 (0.000)*
15.9 (0.014)
12.6 (0.001)*
13.8 (0.007)*
9.7 (0.163)
15.9 (0.000)*
Co-conditions for renal deaths
Aboriginal % (p) Non-Aboriginal % (p)
34.9 (0.000)*
23.0 (0.000)*
25.5 (0.003)*
35.3 (0.000)*
11.6 (0.964)
10.7 (0.688)
8.9 (0.872)
8.2 (0.544)
Co-conditions for stroke deaths
Aboriginal % (p)
Non-Aboriginal % (p)
Diabetes
22.1 (0.001)*
11.9 (0.007)*
Renal
IHD
15.4 (0.872)
13.0 (0.001)*
6.3 (0.544)
18.8 (0.042)
COPD
9.0 (0.158)
9.7 (0.311)
Note: * Significant at P<0.0127 as per Bonferroni method to adjust for multiple comparisons.
15
Table 4 Age-adjusted mortality rate ratios (MRR) for mortality rates by disease, in NT Aboriginal
population compared with NT non-Aboriginal population and all Australia , 1997–2004.
NT Aboriginal to NT non-
NT non- Aboriginal to
NT Aboriginal to Australia
Aboriginal
Australia
8.4 (7.1,10.0)
5.0 (4.1,6.0)
1.0 (1.0,1.4)
13.0 (11.3,14.9)
9.0 (7.4,11.1)
1.3 (1.1,1.5)
COPD
10.3 (8.6,12.5)
5.3 (4.2,6.6)
1.7 (1.4,2.0)
Renal
10.9 (9.2,12.9)
10.3 (8.3,12.8)
1.0(0.8,1.2)
4.3 (3.5,5.1)
4.1 (3.3,5.0)
1.0 (0.8,1.2)
Conditions
IHD
Diabetes
Stroke
Table 5 Annual changes in NT Aboriginal and Australian mortality rates, for five chronic
diseases, by gender, 1997–2004.
Condition
IHD
NT Indigenous male
NT Indigenous
female
Australia male
Australia female
1.3 (-5.0,8.1)
2.9 (-5.2,11.7)
-3.5 (-7.2,0.2)
-3.2 (-7.6,1.4)
-3.6 (-8.8,1.9)
-3.5 (-9.4,2.9)
-0.6 (-1.6,0.3)
-1.3 (-4.2,1.7)
-11.7 (-19.5, -3.1)
-7.1 (-15.8,2.4)
-2.7 (-7.7,2.6)
-4.6 (-9.7,0.7)
-1.0 (-9.6,8.4)
-6.1 (-12.3,0.6)
-5.0 (-10.2,0.5)
-2.8 (-4.5,1.0)
-1.3 (-9.4,7.6)
4.4 (-3.7,13.2)
-4.5 (-9.4,7.6)
-2.0 (-7.2,3.5)
Diabetes
Stroke
COPD
Renal failure
16
17