Tobacco control measures have two main aims – to prevent people ever starting to smoke, and to help those who do to make and maintain a quit attempt. Whilst we know that current smoking presents an obvious risk, and therefore an intervention opportunity, there is less focus on people with a history of smoking. We have well established evidence to demonstrate the risks of any smoking compared to never smoking, and we know that the longer we remain smoke free after being a smoker, the better our health outcomes are (Galucci et al., 2020, Duncan et al., 2019). But what about the population of people who are at a higher risk of physical health conditions relating to their mental illness?

This French study used a large cohort of people with bipolar disorder, whose anonymised data forms part of a national database, FACE-BD. Clients are included in FACE-BD once they have received a specialist clinical assessment and a diagnosis of bipolar disorder is confirmed. Validated tools are used to build a broad profile of people living with bipolar disorder over a long period of time.

Although we often hear about risk of metabolic disorders, high smoking prevalence and poor physical health in relation to people living with schizophrenia, people with bipolar disorders are at a similar high risk of poor physical health and are often overlooked in research.

The current study set out to identify the risk factors associated with being a current, former or never smoker among those with bipolar disorder (Nobile et al, 2023). The authors anticipated that identifying specific risk factors relating to being a never smoker, a former smoker or a current smoker may shed light on specific potential benefits of quitting smoking among those with bipolar disorder; these benefits might then provide content including motivating factors, to use in individualised interventions for people with bipolar disorder who currently smoke.

Methods

The authors used the FACE-BD database to extract clinical information relating to current mental state including key features relating to bipolar disorder and physical health conditions including presence of a metabolic syndrome. They then stratified every case where smoking status information was available into three groups:

1. Current (defined as smoking at least 5 cigarettes per day for at least three months)
2. Former (someone who had smoked more than 100 cigarettes in a lifetime but not in the past year)
3. Never (smoked less than 100 cigarettes in a lifetime).

To compare the association between independent variables, odds ratios were calculated, using ‘never smoker’ as the reference point. This allows the reader to understand whether there is any increased likelihood of each sociodemographic or clinical variable and being a former or current smoker. The authors then conducted a multivariate regression analysis to identify associations between factors.

Results

Data were available for 3,625 people, and these were then stratified into three groups; never smokers = 1,529, former smokers = 416 and current smokers = 1,680. Statistically significant findings from the demographic variables indicated that former smokers were slightly older than current or never smokers, that current and former smokers were less likely to be employed than never smokers, and that former and never smokers were more likely to be in a relationship compared to current smokers.

Among those with bipolar disorder, there was a striking increase in risk of suicide among current smokers compared to never and former smokers, and current smoking was also associated with higher scores for anxiety, depression and emotional lability compared to never smokers, although there was no difference compared to former smokers. Likelihood of a co-morbid substance use disorder increased across never to former and current smokers.

The authors were interested in the metabolic profile of the cohort; and interestingly former smokers showed significantly higher BMI and waist circumference, and more frequent dyslipidemia (e.g., unbalanced or unhealthy cholesterol levels) and metabolic syndrome, in comparison to current and never smokers. Current smokers were more likely to be prescribed antipsychotic medications compared to both former and never smokers.

The multivariate analysis indicated that age (i.e. a younger age of current smoker in comparison with the other two groups) and alcohol and cannabis use disorder were the most significant factors predicting current smoking; current smokers were more likely to have a lifetime history of one or the other of these substance use disorders.

Conclusions

The study found a high prevalence of smoking among people with bipolar disorder compared to the general population, and low quit rates. Current and former smoking was associated with a diagnosis of bipolar disorder type 1; being single; meeting criteria for metabolic syndrome, and experiences of childhood trauma. A comorbid substance use disorder was associated with current smoking and to a lesser extent former smoking. The authors stress that the associations identified are not causal and that prospective studies would be required to demonstrate such a link. However, they conclude that the additional risk factors found among current and former smokers suggest that smoking cessation treatment should remain a high priority for people with a bipolar disorder diagnosis.

Strengths and limitations

A key strength of this study is the large sample size, and wide range of clinical information available. Of note, this study examines combustible tobacco use – the authors have counted users of other ‘non-medical nicotine’ as non or never smokers, although it would be useful to know more about the distinction between medical vs non-medical nicotine. The authors tested for associations across a range of both mental and physical symptoms, and the FACE-BD database offers a comprehensive and representative sample of people living with bipolar disorder in France. The authors took advantage of the sample size to examine former smokers and risk factors in comparison to both never and current smokers.

The authors criticise other smoking studies – which often conflate ‘never smokers’ with ‘former smokers’, and compare these with current smokers, resulting in a lack of detail on potential risk factors among former smokers. To address this, the authors stratified all participants into three groups – although the distinction between these three groups poses a challenge. Current smokers were defined as smoking at least 5 cigarettes per day for the last three months, and former as those who smoked at least one year ago. This distinction leaves a gap – what about someone who smokes less than 5 per day but has done for the past year, or someone who stopped smoking after many years, but only 6 months ago – how did the authors decide which group these examples would fall into? Decisions like this are always required in research, and it is usually helpful to have fewer groups for purposes of comparison, but this does have implications for the findings which relate to the true differences between current and former smokers. This is a challenge across all tobacco research – how to capture, and interrogate the amount someone smokes, or has smoked, and what difference this might make to their latent risks.

Interestingly, the authors have also indicated that light smokers were most often classified as ‘non-smokers’ – and there is prior literature to support the classification of never use as someone who has smoked less than 100 cigarettes in their lifetime (Pomerleau et al., 2004). However it isn’t clear if in this study someone who is a current ‘light smoker’ might be considered a non-smoker, whereas someone who was previously a ‘heavy smoker’ but quit a year ago is a former smoker – the conflation of current smokers (however ‘light’, we know that even one cigarette can have an impact on risk of coronary heart disease (Hackshaw et al., 2018)) and never smokers is not clear.

The potential for overlap between the current and former smokers makes it harder to be certain about the outcomes in this study; and although the authors set out to explore more about the former smoker groups, this has highlighted the challenges in measurement of health behaviours, and the limitations of secondary data use.

Implications for research

We know that the more years spent not smoking, the better our health outcomes, but what is the difference in terms of ongoing risk to health for someone who smoked 30 cigarettes per day for 20 years, but quit ten years ago, compared to someone who smoked five cigarettes per day for 20 years, but hasn’t quit? This isn’t a new issue for tobacco and nicotine researchers. Asking someone if they smoke seems straightforward – the answer would be yes or no, right? But – someone who smokes occasionally, maybe last month when they were at a party, and might feasibly again next week, but don’t have any cigarettes in their bag right now – are they a smoker? Would they say they were if asked? Consider someone was a smoker for 20 years but quit last week – and someone who has never smoked a single puff – both now non-smokers but will have likely very different risk profiles.

We have a set of standards – ‘Russell standard’, (named after Mike Russell, one of the first tobacco researchers  (McNeill & Robson, 2018)) – to help with precisely this question (Piper et al., 2019; West et al., 2005). These standards describe how we attribute smoking status within a smoking cessation trial and help to clarify who could be considered to have quit; they ‘allow’ for just five cigarettes between quit date and final data collection point in the trial for someone still to be considered to have successfully quit. In a clinical setting, carbon monoxide testing (e.g. a ‘breathalyser’ device) can be used to provide a reading which is a reliable indication of current smoking status  (NICE, 2021).

Implications for practice

Our clinical systems may be designed to alert us to screen for and then address current smoking – but we may be missing latent processes among former smokers who have been exposed to the risks associated with smoking which pose a risk, albeit at a diminishing rate over time. The study points to the importance of screening, regular screening, and building a picture of historical risk behaviours as well as current. The harms derived from smoking do not always disappear, although they do reduce over time after quitting (Thomson et al., 2022) .

In clinical settings, the focus of care is often on the prevailing symptoms, although this can mean that people with a mental health diagnosis miss out on physical health interventions in comparison to those without. ‘Parity of esteem’ means that the increased risk factors for people with a diagnosis such as bipolar disorder including health behaviours such as current or past smoking, should not be overlooked, but addressed alongside someone’s mental health (Mitchell et al., 2017). Such action is crucial to reducing the health inequalities, such as the mortality gap, experienced by people with mental health conditions.

Perhaps one of the most significant risk factors found in this study, though perhaps unsurprising, is that smoking is very common among people with a co-morbid substance use disorder alongside their diagnosis of mental health condition. This highlights the complexity of clinical challenges within this group, and the overlapping (or even conflicting) clinical priorities are apparent (Huddlestone et al., 2022). Interventions aimed at addressing smoking within each of these populations are available, but much work remains to increase the quit rates for people with a mental illness as well as substance use disorder. This remains a key factor in the inequalities faced by this group, and interventions which address the convergence of both conditions are urgently required.

Statement of interests

None.

Links

Primary paper

Nobile, B.et al (2023). Physical and mental health status of former smokers and non-smokers patients with bipolar disorder. Acta psychiatrica Scandinavica, 147(4), 373–388. https://doi.org/10.1111/acps.13535

References

Duncan, M. S., Freiberg, M. S., Greevy, R. A., Kundu, S., Vasan, R. S., & Tindle, H. A. (2019). Association of smoking cessation with subsequent risk of cardiovascular disease. Jama, 322(7), 642-650.

Gallucci, G., Tartarone, A., Lerose, R., Lalinga, A. V., & Capobianco, A. M. (2020). Cardiovascular risk of smoking and benefits of smoking cessation. Journal of thoracic disease, 12(7), 3866.

Hackshaw, A., Morris, J. K., Boniface, S., Tang, J., & Milenković, D. (2018). Low cigarette consumption and risk of coronary heart disease and stroke: meta-analysis of 141 cohort studies in 55 study reports. Bmj, 360

Huddlestone, L., Shoesmith, E., Pervin, J., Lorencatto, F., Watson, J., & Ratschen, E. (2022). A Systematic Review of Mental Health Professionals, Patients, and Carers’ Perceived Barriers and Enablers to Supporting Smoking Cessation in Mental Health Settings. Nicotine & Tobacco Research, 24(7), 945-954. 10.1093/ntr/ntac004

McNeill, A., & Robson, D. (2018). A man before his time: Russell’s insights into nicotine, smoking, treatment and curbing the smoking problem. Addiction (Abingdon, England), 113(4), 759-763. 10.1111/add.14043

Mitchell, A. J., Hardy, S., & Shiers, D. (2017). Parity of esteem: Addressing the inequalities between mental and physical healthcare. BJPsych Advances, 23(3), 196-205. 10.1192/apt.bp.114.014266

NICE. (2021). Tobacco: preventing uptake, promoting quitting and treating dependence. National Institute for Health and Care Excellence.

Piper, M. E., Bullen, C., Krishnan-Sarin, S., Rigotti, N. A., Steinberg, M. L., Streck, J. M., & Joseph, A. M. (2019). Defining and Measuring Abstinence in Clinical Trials of Smoking Cessation Interventions: An Updated Review. Nicotine & Tobacco Research, 22(7), 1098-1106. 10.1093/ntr/ntz110

Pomerleau, C. S., Pomerleau, O. F., Snedecor, S. M., & Mehringer, A. M. (2004). Defining a never-smoker: Results from the nonsmokers survey. Addictive Behaviors, 29(6), 1149-1154. 10.1016/j.addbeh.2004.03.008

Thomson, B., Emberson, J., Lacey, B., Lewington, S., Peto, R., Jemal, A., & Islami, F. (2022). Association Between Smoking, Smoking Cessation, and Mortality by Race, Ethnicity, and Sex Among US Adults. JAMA Network Open, 5(10), e2231480. 10.1001/jamanetworkopen.2022.31480

West, R., Hajek, P., Stead, L., & Stapleton, J. (2005). Outcome criteria in smoking cessation trials: proposal for a common standard. Addiction, 100, 299-303. 10.1111/j.1360-0443.2005.00995.x

Photo links

• Photo by Pawel Czerwinski on Unsplash
• Photo by Jungwoo Hong on Unsplash
• Photo by Jack Anstey on Unsplash