Abstract
Professor Chapman suggests we overstated the contribution of mental illness to smoking when we reported that 32.4% of current smokers have a mental illness because we included all smoking not just daily smoking [1,2]. We have calculated the figures for daily smoking, and the proportions are very similar: of 2 921 000 daily smokers, 951 000 had a mental disorder in the past 12 months, some 32.6%.
Chapman wonders why the prevalence of mental health problems has increased substantially between the 2004 figures he cites and the 2007 we reported. The 2004 and 2007 surveys he cites measure very different things. There is no evidence of any substantive change in prevalence of mental disorders between 1997 and 2007 [3]. While the 2007 National Survey of Mental Health and Wellbeing (NSMHWB) assessed mental disorders according to ICD-10 and DSM-IV criteria based on a detailed structured interview, the 2004 National Health Survey (NHS) did not specifically assess mental disorders but asked respondents to self-report if they had any mental or behavioural problems. The majority of those assessed with mental disorders in the NSMHWB did not access or perceive the need to access health services for their disorders. Perhaps many of this group would be unlikely to self-report mental or behavioural problems in the NHS. Professor Chapman speculates that those identified in the NSMHWB but not in the NHS would have less severe mental disorders that would not impact on their smoking status. The NSMHWB data, however, do not support this. An estimated 64% of people with 12 month mental disorders do not use any services for mental health problems. Of the 2 092 000 people with mental disorders who do not use services, some 740 000 are current smokers (35.4%). Of the 1 106 000 people with mental disorders who do access a medical service 416 600 are current smokers (37.6%). Both smoking rates are substantially higher than the 18.8% of Australian adults with no 12 month mental disorder.
The diagnostic category ‘harmful alcohol consumption’ Professor Chapman refers to from the NSMHWB is not related to NHMRC criteria for high-risk or risky drinking. It is based on ICD-10 criteria that require functional impairment due to dependence on or abuse of alcohol and thus the prevalence of this disorder is lower than that in surveys that assess high-risk or risky drinking status.
Professor Chapman claims we unfairly criticized the National Preventative Health Taskforce (NPHT) technical report [4], of which Professor Chapman was a co-author, for its lack of attention to the issue of mental illness and smoking because he claims two recommendations and a section of the report deal with this issue. The NPHT report and recommendations relate only to smoking cessation support within mental health services. The majority of people with mental disorders who do not access services will not benefit from service-based smoking cessation efforts. The only mention of mental illness outside the context of psychiatric services in the NPHT report was to suggest that people with common problems such as anxiety and depression would benefit equally as anyone else from the broad population health measures used in tobacco control (p. 43). This is contrary to the evidence. The 2007 NSMHWB data suggest that people with common mental illness are more likely to take up smoking and less likely to quit, thus smoking on average for longer and presumably suffering greater health consequences as a result. We are pleased that the recently released National Preventative Health Strategy [5] has recognized the broader implications of mental illness outside the hospital setting for tobacco control.
Tobacco control groups have traditionally preferred to see smoking as a single issue. Not surprisingly they have had their best success in people for whom smoking is the single major health issue they face. This group now have very low smoking rates. While this needs to be maintained, a more nuanced approach to population health smoking cessation could aim to specifically target the significant proportion of smokers who are simultaneously dealing with other complex issues in their lives. Professor Chapman has previously suggested that money spent on targeted anti-smoking strategies is money taken away from the more effective broad population-wide campaign and thus money poorly spent [6]. A focus on the single issue campaign, however, which is most effective in one subgroup of the population, has the side-effect of increasing inequality between those with and without mental illness, while providing the least support to a segment of the community who already suffer considerable disadvantage, and whose mental illness may require more than average levels of support to achieve good health outcomes.