Smoking and Mental Illness: Results From Population Surveys in Australia and the United States

David Lawrence; Francis Mitrou; Stephen R Zubrick


BMC Public Health. 2009;9:285 

In This Article

Abstract and Background


Background: Smoking has been associated with a range of mental disorders including schizophrenia, anxiety disorders and depression. People with mental illness have high rates of morbidity and mortality from smoking related illnesses such as cardiovascular disease, respiratory diseases and cancer. As many people who meet diagnostic criteria for mental disorders do not seek treatment for these conditions, we sought to investigate the relationship between mental illness and smoking in recent population-wide surveys.
Methods: Survey data from the US National Comorbidity Survey-Replication conducted in 2001–2003, the 2007 Australian Survey of Mental Health and Wellbeing, and the 2007 US National Health Interview Survey were used to investigate the relationship between current smoking, ICD-10 mental disorders and non-specific psychological distress. Population weighted estimates of smoking rates by disorder, and mental disorder rates by smoking status were calculated.
Results: In both the US and Australia, adults who met ICD-10 criteria for mental disorders in the 12 months prior to the survey smoked at almost twice the rate of adults without mental disorders. While approximately 20% of the adult population had 12-month mental disorders, among adult smokers approximately one-third had a 12-month mental disorder–31.7% in the US (95% CI: 29.5%–33.8%) and 32.4% in Australia (95% CI: 29.5%–35.3%). Female smokers had higher rates of mental disorders than male smokers, and younger smokers had considerably higher rates than older smokers. The majority of mentally ill smokers were not in contact with mental health services, but their rate of smoking was not different from that of mentally ill smokers who had accessed services for their mental health problem. Smokers with high levels of psychological distress smoked a higher average number of cigarettes per day.
Conclusion: Mental illness is associated with both higher rates of smoking and higher levels of smoking among smokers. Further, a significant proportion of smokers have mental illness. Strategies that address smoking in mental illness, and mental illness among smokers would seem to be important directions for tobacco control. As the majority of smokers with mental illness are not in contact with mental health services for their condition, strategies to address mental illness should be included as part of population health-based mental health and tobacco control efforts.


Smoking remains one of the leading causes of preventable disease and death both in Australia and the United States.[1,2] Efforts to reduce the prevalence of smoking continue to be of public health significance. The association between smoking and mental illness has been known and studied for many years, but the focus of much of the work in this field has been on people with severe mental illness, such as schizophrenia, or people being treated by psychiatric services.[3–8] For instance, a recent meta-analysis by de Leon and Diaz of 42 individual studies reported that people with schizophrenia had odds of smoking 5.3 (95% CI: 4.9–5.7) times higher than the general population.[9]

Glassman et al reported an association between smoking and major depression from the St Louis Epidemiological Catchment Area Survey.[10] Lasser et al reported the first major population study in the US showing the substantial proportion of smokers who met DSM-III-R criteria for mental disorders, many of whom were not in contact with mental health services.[11,12] They reported that people with DSM-III-R mental illnesses in the month prior to the survey had a smoking rate twice as high as people with no mental illness, and consumed an estimated 44% of cigarettes smoked by adults in the United States.

The excess mortality among people with mental illness due to common conditions for which smoking is a known risk factor, such as cardiovascular disease, respiratory disease and cancers, has been extensively described, from both service-based and community-based samples, and there is evidence of a growing disparity in mortality rates between those with and without mental illness.[13–17] For example, in Western Australia, cardiovascular mortality fell significantly in the general population between 1980–1998, but there was no decline among people with mental disorders.[18] One possible explanation for this phenomenon may be that the public health interventions that have helped to reduce cardiovascular mortality in the general community, such as the efforts that have reduced the prevalence of smoking in both Australia and the US, may have been less effective among people who have mental illness.

It has been hypothesised that one reason for the strong association between mental illness and current smoking is that mental illness is a factor in smoking initiation. Depression and anxiety in teenagers have been found to be strong predictors of smoking experimentation and the transition to daily smoking.[19–21] However, smoking has also been associated with the onset of psychiatric symptoms in teenagers.[21,22] This has suggested that smoking and depressive and/or anxiety symptoms may have onset around the same time, possibly associated with common prior causes.[23–25] However, controlling for common causal factors does not completely remove the predictive ability of smoking on mental illness, particularly depression and anxiety and vice versa. This suggests the possibility that separate causal mechanisms may operate in both directions, in addition to the predictive ability of common causal factors.[21,26–28]

The large overlap between mental illness and smoking is not entirely surprising considering the known effects of nicotine on the brain. Nicotine is a psychostimulant that effects several neuroregulators that influence behaviour and mood.[29,30] In some circumstances, nicotine can relieve symptoms of both depression and anxiety.[31,32] Nicotine cessation can also precipitate depressive symptoms, particularly in people with a history of major depression.[33,34] However, it has been questioned whether these are independent depressive symptoms or unpleasant withdrawal symptoms.[35,36] The onset of depressive symptoms following smoking cessation has been linked with lower quit rates, and most commonly occurs in people with depression.[37–39] These factors have lead to the self-medication hypothesis–that smokers with mental illness choose to smoke because it is the easiest, most readily accessible way to control symptoms of mental illness, especially for those who are not receiving any prescribed form of treatment for their mental health condition.[40]

While the self-medication theory implies there is a therapeutic benefit to smoking which people with depressive or anxiety symptoms find helpful, research suggests that smoking provides temporary relief from immediate symptoms, while overall creating a greater level of anxiety and stress.[31,35,41] Thus the alleviation of stress and anxiety by smoking may be part of the withdrawal feedback mechanism. It has been suggested that the self-medication hypothesis has too often been used as a justification for not acting to curb cigarette smoking in this population despite the fact that nicotine is not regarded as the most appropriate therapy for any mental health problem, and that even if it were, cigarettes would not be an appropriate form of administering it.[37]

The study of Lasser et al was based on data from the US National Comorbidity Survey which was collected in 1992. Smoking rates have been in decline in the US since that time.[42] More current data are now available with the release of the public use file from the US National Comorbidity Study-Replication which was conducted between 2001 and 2003, and data from the 2007 Australian Survey of Mental Health and Wellbeing. We hypothesised that despite the decline in overall smoking rates in both the US and Australia since 1992, people with mental illness would continue to represent a disproportionally high number of smokers.

We used data from two recent nationally representative surveys that employed the WHO Composite International Diagnostic Interview (CIDI) to estimate prevalence of mental health disorders among smokers–the National Comorbidity Study-Replication in the US and the 2007 Australian Survey of Mental Health and Wellbeing. We also examined data from the annual US National Health Interview Survey, which has used a short measure of psychological distress as a proxy of severe mental illness, to examine whether it would be feasible to monitor the relationship between mental illness and smoking as part of routine surveillance systems.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.