NDARC Technical Report No. 80 (1999)


Tobacco is one of the most widely used psychoactive substances in the general population. The majority of Australian adults report having used tobacco at some point in their lives; and although many people report regular use, few perceive tobacco as a problematic drug.

There is considerable clinical evidence to suggest that tobacco use is more prevalent among persons in treatment for anxiety disorders, depression, and among persons with schizophrenia. Furthermore, there is evidence to suggest that persons with substance use problems are even more likely to smoke.

However, while these data suggest that people coming to the attention of treatment services have higher rates of smoking than the general population, clinical samples are subject to a number of selection biases. This means that although a relationship might exist in treatment populations, it may not be representative of the relationship in the general population. Although there has been some examination of the links between anxiety, depression and smoking using community surveys in the US, there has been no detailed examination of the relationship between tobacco use and mental health problems in the Australian population. Furthermore, there has been no epidemiological research examining the relationship between tobacco use and other substance use problems.

This report examined the associations in the Australian population between tobacco use, other drug use problems, and mental health problems. This was made possible using data from the 1997 National Survey of Mental Health and Well-Being, which was a representative sample of 10,641 Australian adults aged 18 years and over. The interview assessed tobacco use, DSM-IV anxiety, affective and substance use disorders, and screened for psychosis. There was also assessment of psychological distress and disability.

In addition to examining the simple associations between mental health problems, drug use problems, and smoking, a number of demographic variables were examined, as well as a measure of personality, the neuroticism scale of the EPQ. These variables were considered in a multivariate analysis, in addition to other drug use, to examine the possibility that univariate patterns were simply due to associated factors.

One in four persons (25%) reported current smoking. Males were slightly more likely to smoke than females (27% vs. 23%). Those who were younger were also much more likely to smoke, with 34% of 18-24 year olds currently smoking, compared to 8% of those aged 75 years and older.

Smoking was also related to a number of indicators of socioeconomic status. Those with less formal education were more likely to smoke, as were those who were unemployed, and had greater social disadvantage. Finally, those with higher levels of neuroticism (i.e. higher levels of moodiness, anxiety and irritability) were also more likely to be smokers.

Anxiety and affective disorders were more than twice as common among smokers, with around one in 10 smokers having an anxiety disorder (9.3%) and one in 10 (10%) having an affective disorder. In comparison, only around one in 20 non-smokers had an anxiety (4.5%) or an affective (5.4%) disorder. Those who reported current smoking were around 5 times more likely to screen positively for psychosis (95%CI 3.2, 6.8). Smokers also reported significantly higher levels of psychological distress and disability due to emotional problems than non-smokers. These significant differences all remained even after the effects of demographic factors, neuroticism and other drug use were taken into account.

Substance use disorders were even more strongly related to smoking. Compared to non-smokers, current smokers were around 4 times more likely to have an alcohol use disorder (95%CI 3.4, 4.7), around 8 times more likely to have a cannabis use disorder (95%CI 5.7, 10.2), and 5 times more likely to have another drug use disorder (sedative, stimulant or opiate; 95%CI 3.1, 7.1). Again, these relationships remained significant even after accounting for demographics, neuroticism and other drug use.

These findings indicated that tobacco use is strongly related to mental health problems (as assessed with symptom measures or as mental disorders) and other substance use problems. These relationships persisted after accounting for a range of confounding variables.

Research has found that mental health problems reduce the likelihood of quitting smoking. Hence, it appears that there is a need to further examine interventions for smokers with mental health and substance use problems. Furthermore, general practitioners and health professionals need to consider the possibility that smokers may have a number of other problems that will decrease their chances of successfully giving up smoking.



Louisa Degenhardt, Wayne Hall
Date Commenced
03 Dec 1999
Resource Type
Technical Reports