Aboriginal and Torres Strait Islander Peoples: Smoking Trends, Australia

This publication describes trends in smoking prevalence, smoking initiation and smoking cessation (quitting) among the Aboriginal and Torres Strait Islander population during the 20 year period from 1994 to 2014–15. Comparisons with trends for the non-Indigenous population during the period 1995 to 2014–15 are included where relevant. The key findings are as follows.

• Smoking prevalence decreased among Aboriginal and Torres Strait Islander peoples aged 18 years and over from 1994 to 2014–15.
• Smoking initiation, as indicated by smoking prevalence among Aboriginal and Torres Strait Islander peoples aged 15–17 years, also decreased over the same period. The annual decrease was faster for the period 2008 to 2014–15 than for the period 1994 to 2004–05, which suggests it is possible that increasing attention to and specific funding for Aboriginal and Torres Strait Islander tobacco control since 2008 is having an impact.
• Successful smoking cessation (quitting) increased among Aboriginal and Torres Strait Islander peoples aged 18 years and over from 2002 to 2014–15, as indicated by increases in the ratio of the number of adults who had successfully quit smoking as a proportion of the number of adults who had ever smoked. Increases were faster for Aboriginal and Torres Strait Islander peoples living in non-remote areas than in remote areas.
• There appears to have been no change to the gap between smoking prevalence in the Aboriginal and Torres Strait Islander adult population and the non-Indigenous adult population during the 20 years to 2014–15, which was around 30 percentage points.

The Australian Bureau of Statistics (ABS) acknowledges and thanks the Menzies School of Health Research for their contribution to this publication. This publication is an outcome of a joint project to examine trends in smoking by Aboriginal and Torres Strait Islander peoples over a 20 year period using regression modelling and data from 12 health and social surveys by the ABS. It helps fill a gap in information about trends in smoking habits by using data from all surveys, rather than comparing prevalence at two points in time, and also facilitates an examination of whether the gap in smoking rates for the Aboriginal and Torres Strait Islander population and the non-Indigenous population is narrowing in a significant way. The partnership with Menzies has involved an exchange of knowledge and methods, as well as joint analysis and interpretation of the data, and resulted in a mutual understanding of smoking behaviours and the social environment impacting on those behaviours.

Tobacco smoking causes many health problems including chronic diseases such as ischaemic heart disease, stroke, chronic obstructive pulmonary disease, asthma, diabetes and various cancers (for example, lung, throat, liver and colorectal cancers).¹ Smoking is estimated to be responsible for 23% of the health gap between Aboriginal and Torres Strait Islander peoples and other Australians.²

Since 2008, there has been a greater focus on reducing smoking prevalence among Aboriginal and Torres Strait Islander peoples and increased funding for tobacco control programs for them. For example, the Australian Government funded the Indigenous Tobacco Control Initiative for $14.5 million over four years from 2008–09 to 2011–12. That initiative was replaced before the end of its term by the Tackling Indigenous Smoking and Healthy Lifestyle program for $100.6 million over four years from 2010–11 to 2013–14 as part of its Indigenous Chronic Disease Package. A review of that program in late 2014 led to a redesigned Tackling Indigenous Smoking program, funded for $116.8 million over three years from 2015–16 to 2017–18.

In addition to these programs, the Australian Government has invested in social marketing campaigns focused on Aboriginal and Torres Strait Islander peoples who smoke, with the national 'Break the Chain' campaign launched in 2010 and the 'Don’t Make Smokes Your Story' campaign launched in 2016.

State and territory governments also have tobacco control programs. For example, in Western Australia, the Country Health Service are developing and implementing multifaceted and culturally appropriate tobacco control projects in a number of regions jointly funded by the Western Australian Government and the Australian Government. Aboriginal community controlled health services also run many tobacco control activities, including local campaigns to raise awareness and support quitting, such as 'Rewrite Your Story' run by Nunkuwarrin Yunti of South Australia Inc.

The Australian Bureau of Statistics (ABS) has collected information about the smoking habits of Aboriginal and Torres Strait Islander peoples since 1994. This includes information about smoking prevalence, smoking initiation and smoking cessation (quitting). These information collections are used to monitor the collective impact of programs such as those described above plus many other local initiatives.

A total of 12 surveys were available for analysis – six for the Aboriginal and Torres Strait Islander population and six for the non-Indigenous population. The six surveys for the Aboriginal and Torres Strait Islander population were:

• 1994 National Aboriginal and Torres Strait Islander Survey (NATSIS),
• 2002, 2008, 2014–15 National Aboriginal and Torres Strait Islander Social Surveys (NATSISS),
• 2004–05 National Aboriginal and Torres Strait Islander Health Survey (NATSIHS), and
• 2012–13 Australian Aboriginal and Torres Strait Islander Health Survey (AATSIHS).
   

The six surveys for the non-Indigenous population were:

• 1995, 2001, 2004–05 and 2007–08, 2014–15 National Health Surveys (NHS), and
• 2011–12 Australian Health Survey (AHS).
   

These surveys were large, national, multistage random household surveys. The non-Indigenous surveys excluded Very Remote areas of Australia and discrete Aboriginal and Torres Strait Islander communities but these exclusions are unlikely to affect national estimates for the non-Indigenous population. Non-private dwellings (for example, hospitals, hostels, caravan parks and prisons) were also excluded from all surveys. Results were weighted to the estimated resident population.

All surveys used similar methods and questions with only minor variations. Trends for smoking prevalence and cessation are reported for people aged 18 years and over, and smoking prevalence for people aged 15 to 17 years are reported as an indicator of trends in smoking initiation. Further information about which smoking questions were asked in each survey is available in Appendix 1 of the Methodology: Smoking Questions by Survey.

Due to the irregular nature of the surveys involved for both the Aboriginal and Torres Strait Islander and non-Indigenous populations, the ABS has used regression modelling to examine trends in estimates of average annual relative change and absolute change between 1994 and 2014–15 for the Aboriginal and Torres Strait Islander population and between 1995 and 2014–15 for the non-Indigenous population.

• Relative change measures the percentage change between two proportions. For example, if smoking prevalence in population A decreased from 50% to 35% and in population B from 30% to 20%, then the relative change in population A was a 30% decrease and in population B was a 33% decrease. This tells us that the prevalence rates in both populations have fallen, but does not tell us anything about the gap between them.
• Absolute change describes a change between two proportions as a percentage point difference. In the example above, population A has experienced an absolute decrease of 15 percentage points while population B had an absolute change of 10 percentage points. This tells us that the gap between these two populations is changing, as population A has had a larger percentage point decrease than population B. Furthermore knowing that population A is larger than population B tells us that the gap is in fact narrowing.
   

This publication explores:

• the average annual rate of change in smoking prevalence, initiation and cessation for the Aboriginal and Torres Strait Islander population during the 20 year period from 1994 to 2014–15,
• whether the rate of change has increased since the greater focus on and funding for reducing smoking prevalence among the Aboriginal and Torres Strait Islander population since 2008, and
• whether the gap between smoking prevalence, initiation and cessation among the Aboriginal and Torres Strait Islander population and the non-Indigenous population has changed significantly during the 20 years to 2014–15, by comparing the average annual absolute change in the two populations.
   

For the relative change analysis, there were more than 170,000 individual data points from respondents aged 18 years and over across the surveys available for analysis. The number of data points available for analysis for estimates of smoking prevalence for 15–17 year olds was much smaller (less than 7,000). For the absolute change analysis, estimates of smoking prevalence by age-sex-remoteness category were analysed for each survey, which gave a maximum of 144 data points for all of the 12 surveys.

The analysis of average annual absolute change tells a different story and provides slightly different conclusions than the more typical statistical analysis of trends using average annual relative changes. This is partly because they are measuring different concepts, and partly because the power to detect a significant effect for average annual absolute change is less than that for average annual relative change due to the use of aggregated data, with fewer data points available for analysis, rather than individual data.

As a person’s smoking habits can be affected by their age, sex, where they live and their level of education, estimates of the relative and absolute change were produced controlling for some or all of these variables.

• Estimates of the average annual relative change for smoking prevalence and smoking cessation were produced controlling for age, sex, remoteness, state/territory and/or education (whether they had completed Year 12 or equivalent or not).
• Estimates of the average annual relative change for smoking initiation among people aged 15–17 years were produced controlling for the same variables except education.
• Estimates of the average annual absolute change were produced controlling for age, sex and remoteness only. Additional variables like state/territory and education are not controlled for due to low accuracy of estimates for small cells.
   

Further information about the regression modelling undertaken, its limitations and the variables used is available in the Explanatory Notes and Technical Note: Modelling Average Annual Change of the Methodology.

According to the 2016 ABS Census of Population and Housing 649,171 people identified as being of Aboriginal and/or Torres Strait Islander origin. Of these, 49.6% were male and 50.4% were female. Aboriginal and Torres Strait Islander peoples represented 3% of the total Australian population.

In the 2016 Census, the Aboriginal and Torres Strait Islander population had a younger age distribution than the non-Indigenous population, reflecting higher fertility and lower life expectancy. In 2016, the median age of the Aboriginal and Torres Strait Islander population was 23 years, compared with 38 years for the non-Indigenous population.

It is projected that in 2016 around 8 in 10 Aboriginal and Torres Strait Islander peoples lived in non-remote areas of Australia, while the remainder lived in remote areas.

Further information about the population size and structure is available in Census of Population and Housing - Counts of Aboriginal and Torres Strait Islander Australians, 2016 (cat. no. 2075.0) and Estimates and Projections, Aboriginal and Torres Strait Islander Australians, 2001 to 2026 (cat. no. 3238.0).

All changes described in the following sections are statistically significant (p<0.05) unless otherwise noted.

In 1994, 55% of Aboriginal and Torres Strait Islander peoples aged 18 years and over were smokers; 20 years later, in 2014–15, this had declined to 45%. Over a similar 20 year period, the proportion of non-Indigenous people aged 18 years and over who were smokers also declined, from 24% in 1995 to 16% in 2014–15.

Between 1994 and 2014–15, the proportion of Aboriginal and Torres Strait Islander peoples aged 18 years and over in non-remote areas who were smokers declined from 55% to 42%, while the proportion in remote areas remained relatively stable at between 54% and 56%.

In 2014–15, 45% of the Aboriginal and Torres Strait Islander population aged 18 years and over were smokers. Of these:

• the vast majority (93%) were daily smokers, and
• the remainder (7%) smoked less than daily.
   

One quarter (25%) of Aboriginal and Torres Strait Islander peoples aged 18 years and over were ex-smokers, and almost one third (31%) had never smoked.

Aboriginal and Torres Strait Islander males were more likely than Aboriginal and Torres Strait Islander females to be smokers (47% compared with 42%), and Aboriginal and Torres Strait Islander peoples in remote areas were more likely than those in non-remote areas to be smokers (54% compared with 42%).

In 2014–15, Aboriginal and Torres Strait Islander peoples aged 18 years and over who had completed Year 12 or equivalent were less likely to be smokers than those who had not completed Year 12 or equivalent or who had never attended school (29% compared with 51%).

There were significant average annual relative decreases in smoking prevalence among both the Aboriginal and Torres Strait Islander population and the non-Indigenous population aged 18 years and over during the period 1994/1995 to 2014–15. After controlling for age and sex only, the estimated annual relative decrease was not significantly different among the Aboriginal and Torres Strait Islander population (a decrease of 1.9% per year) compared with the non-Indigenous population (a decrease of 2.5% per year). However, after controlling for all variables (age, sex, remoteness, state/territory and education), the estimated annual relative decrease was significantly different between these two populations, with a slower rate of decrease among the Aboriginal and Torres Strait Islander population (1.0% per year) than among the non-Indigenous population (2.7% per year). Further detail about these results, including significance levels and confidence intervals, can be found in Appendix 2 of the Methodology: Results of Modelling.

Among Aboriginal and Torres Strait Islander peoples aged 18 years and over, the estimated annual relative decreases in smoking prevalence between 1994 and 2014–15 were faster for those in younger age groups, after controlling only for sex. There was a 3.1% decrease per year for those aged 18–24 years, a 2.2% decrease per year for those aged 25–34 years and a 1.9% decrease per year for those aged 35–44 years, while smoking prevalence did not change for those aged 45–54 years and 55 years and over.

Among Aboriginal and Torres Strait Islander peoples aged 18 years and over in non-remote areas, the annual relative decreases in smoking prevalence were significantly faster between 1994 and 2014–15 for those who had completed Year 12 or equivalent (3.3% per year) compared with those had not completed Year 12 or equivalent or who had never attended school (1.3% per year), after controlling for age, sex and state/territory. In contrast, in remote areas, smoking prevalence did not change for those who had completed Year 12 or equivalent, and increased by 1.2% per year for those who had not completed Year 12 or had never attended school.

The estimated annual relative change among Aboriginal and Torres Strait Islander peoples aged 18 years and over decreased significantly faster during the period 2008 to 2014–15 (a decrease of 2.1% per year) than during the period 1994 to 2004–05 (an increase of 0.7% per year), after controlling for all variables (age, sex, remoteness, state/territory and education). There was no such difference among the non-Indigenous population aged 18 years and over when comparing the periods 1995 to 2008 and 2011–12 to 2014–15; however, the second period only includes data from two surveys (2011–12 and 2014–15).

Smoking prevalence for the Aboriginal and Torres Strait Islander population aged 18 years and over was 31 percentage points higher than that for the non-Indigenous population (55% compared with 24%) in 1994/1995, and 29 percentage points higher in 2014–15 (45% compared with 16%). This section explores whether the average annual percentage point change for each population over this period was statistically significant.

The results showed there were significant average annual absolute decreases in smoking prevalence among both the Aboriginal and Torres Strait Islander population and the non-Indigenous population aged 18 years and over between 1994/1995 and 2014–15; however, the decreases were not significantly different between the two populations and so the percentage point gap did not change over this period. The estimated annual absolute decrease in smoking prevalence for those aged 18 years and over was 0.3 percentage points per year for both the Aboriginal and Torres Strait Islander population and the non-Indigenous population, after controlling for all variables (age, sex and remoteness).

Among Aboriginal and Torres Strait Islander peoples aged 18 years and over, the estimated annual absolute decrease in smoking prevalence was 0.6 percentage points per year for those in non-remote areas, compared with no significant change for those in remote areas, after controlling for age and sex only.

The estimated average annual absolute change among Aboriginal and Torres Strait Islander peoples aged 18 years and over during the period 2008 to 2014–15 was no different to that for the period 1994 to 2004–05, after controlling for all variables (age, sex, and remoteness). Similarly, there was no significant difference among the non-Indigenous population aged 18 years and over when comparing the periods 1995 to 2008 and 2011–12 to 2014–15 (decrease of 0.2 percentage points per year for each period). However, as noted earlier in the results for relative change, the second period only includes data from two surveys.

Most of the Aboriginal and Torres Strait Islander population (69% in 2012–13) and non-Indigenous population (58% in 2014–15) aged 18 years and over who have ever smoked daily, commenced smoking before 18 years of age. Young people who do not initiate smoking during this time are unlikely to do so later in life¹, and those who start earlier are more likely to smoke more heavily and less likely to quit as adults². Preventing the onset of adolescent smoking is therefore a key part of any tobacco control strategy to reduce smoking prevalence and the associated morbidity and mortality.

In this publication, information about smoking prevalence for Aboriginal and Torres Strait Islander peoples aged 15–17 years is used as an indicator of trends in smoking initiation. Smoking data for this age group was not collected in 2004–05.

In 1994, 30% of the Aboriginal and Torres Strait Islander population aged 15–17 years were smokers; by 2014–15 this had dropped to 17%. In 2014–15, there was no significant difference between Aboriginal and Torres Strait Islander males and females aged 15–17 years who were smokers (22% and 14% respectively). Similarly, there was no significant difference between Aboriginal and Torres Strait Islander peoples aged 15–17 years who were smokers in non-remote and remote areas (16% and 24% respectively). While these differences appear large the statistical significance of these differences is influenced by the relatively small sample size (when compared with the sample size for 18 years and over), which increases the variability associated with these estimates.

There was a significant average annual relative decrease in smoking prevalence for Aboriginal and Torres Strait Islander peoples aged 15–17 years between 1994 and 2014–15. The estimated annual relative decrease in smoking prevalence for those aged 15–17 years was 4.3% per year, after controlling for age and sex only. After controlling for all variables (age, sex, remoteness and state/territory), the estimated annual relative decrease in smoking prevalence among those aged 15–17 years was similar at 4.1% per year.

The average annual relative decrease in smoking prevalence for those aged 15–17 years did not vary significantly between different categories of age, sex, remoteness and state/territory, after controlling for all these variables.

There was also no significant difference in the estimated annual relative change in smoking prevalence among Aboriginal and Torres Strait Islander peoples aged 15–17 years during the period 2008 to 2014–15 when compared with the period 1994 to 2002, with both periods showing no significant change per year after controlling for all variables (age, sex, remoteness and state/territory). However, the earlier period only includes data from two surveys (1994 and 2002).

Between 1994 and 2014–15, the smoking prevalence rate for Aboriginal and Torres Strait Islander peoples aged 15–17 years decreased by 13 percentage points (from 30% to 17%). This section explores whether the average annual percentage point change over this period was statistically significant.

The average annual absolute percentage point change in smoking prevalence for Aboriginal and Torres Strait Islander peoples aged 15–17 years varied significantly by age, with no significant change per year for those aged 15 years, a decrease of 0.7 percentage points per year for those aged 16 years, and a decrease of 1.7 percentage points per year for those aged 17 years, after controlling for sex and remoteness only.

The estimated annual absolute percentage point change in smoking prevalence among Aboriginal and Torres Strait Islander peoples aged 15–17 years was also significantly faster during the period 2008 to 2014–15 (1.9 percentage point decrease per year) than during the period 1994 to 2002 (no significant change per year) after controlling for all variables (age, sex and remoteness). However, the earlier period only includes data from two surveys (1994 and 2002).

Trends in smoking cessation (quitting) are a function of trends in the proportion of smokers attempting to quit and the proportion of these attempts which are successful. The following sections look at:

• Quit ratio
• Attempts to quit smoking
   

It is encouraging that smoking prevalence among the Aboriginal and Torres Strait Islander population has decreased during the 20 year period from 1994 to 2014–15, with significant relative and absolute annual decreases in smoking prevalence among those aged 18 years and over and in smoking initiation among those aged 15–17 years, accompanied by significant relative and absolute increases in successful quitting for those aged 18 years and over. These trends will lead to rapid and sustained health gains among the increasing numbers of successful quitters, and avert the predictable health consequences of smoking among the increasing numbers of Aboriginal and Torres Strait Islander peoples who are not starting to smoke. While faster declines would lead to greater health gains sooner, these significant improvements demonstrate that improvement is achievable and is happening now.

The slower relative annual decrease in smoking prevalence among the Aboriginal and Torres Strait Islander population (1.0%) compared with the non-Indigenous population (2.7%) aged 18 years and over appears discouraging; however, the average absolute annual decrease in percentage points is the same for both populations (0.3 percentage points). The slower relative decrease despite the same absolute decrease for each population reflects the much higher smoking prevalence at the start of the period among the Aboriginal and Torres Strait Islander population (55%) compared with the non-Indigenous population (24%). The Council of Australian Government’s target for reducing Aboriginal and Torres Strait Islander smoking prevalence refers to halving smoking prevalence among the Aboriginal and Torres Strait Islander population from 2008 within 10 years (by 2018). While a 50% reduction in 10 years now appears unlikely, this publication has demonstrated that significant progress towards this target is being made as evidenced by the significant average annual relative decreases in smoking prevalence for those aged 18 years and over since 2008.

In contrast, most other policy debate about Aboriginal and Torres Strait Islander peoples is about Closing the Gap. Using the size of the gap between smoking prevalence for the Aboriginal and Torres Strait Islander population and the non-Indigenous population as a measure, there has been no significant change in the gap for people aged 18 years and over between the two populations.

Research suggests that it takes about 10 years after implementation of a program before its effects can be measured.¹ The impact of new national tobacco control policies targeted to the Aboriginal and Torres Strait Islander population since 2008 on national smoking prevalence rates may therefore not be apparent until about 2018, so any evidence of an impact on smoking trends before then is very encouraging. This publication found faster absolute annual decreases in smoking prevalence among Aboriginal and Torres Strait Islander peoples aged 15–17 years for the period 2008 to 2014–15 than for the period 1994 to 2002. It is possible that the increasing attention to and specific funding for Aboriginal and Torres Strait Islander tobacco control since 2008 is having an impact on smoking initiation by young Aboriginal and Torres Strait Islander peoples.

Continuing the observed trends of increasing cessation among Aboriginal and Torres Strait Islander peoples will require a comprehensive mix of tobacco control policies and activities. A common theme of many research studies has been the importance of family in motivating Aboriginal and Torres Strait Islander peoples to quit smoking, in addition to other factors common in other settings such as health concerns and cost.² National and local Aboriginal and Torres Strait Islander tobacco control campaigns, such as the national ‘Break the Chain’ campaign, have featured messages about quitting to reduce the impact of smoking on other family members.

Previous ABS publications have shown lower smoking prevalence and greater improvements for Aboriginal and Torres Strait Islander peoples in non-remote areas than remote areas. This publication has also found faster relative and absolute decreases in smoking prevalence and faster relative and absolute increases in successful quitting among Aboriginal and Torres Strait Islander peoples aged 18 years and over in non-remote areas than in remote areas. However, there were greater increases in the proportion of smokers who had tried to quit smoking in the previous 12 months in remote areas than in non-remote areas, which suggests attempts to quit smoking by those in remote areas are not being translated into successful quitting at the same rate as for those in non-remote areas. This may be due to greater exposure and access to cessation support services and tobacco control policies and activities (both mainstream, like smoking restrictions in buildings and public places, and those specifically targeting Aboriginal and Torres Strait Islander peoples), greater exposure to anti-smoking social norms, or broader social determinants of health in non-remote areas.

This release previously used catalogue number 4737.0