Severe mental illness (SMI) (e.g., schizophrenia, bipolar disorder, psychosis, etc) can be characterised as long-term, complex conditions associated with substantial impairments in social, emotional and physical health. Individuals living with SMI are estimated to die 15–20 years earlier than those living without SMI. This reduced life expectancy is driven largely by cardiovascular disease and cancer, both of which are strongly associated with smoking cigarettes (Momen et al., 2022; Callaghan et al., 2014). Around 40% of people living with SMI smoke tobacco and, compared with the general population, they tend to start smoking at a younger age and smoke more heavily (Royal College of Physicians, 2013; Public Health England, 2020).
Reducing or quitting smoking is not only beneficial for the physical health of those with SMI, but there is some research to show that stopping smoking is also associated with reduced depression, anxiety and stress (Taylor et al., 2014; Brose et al., 2011; McDermott et al., 2013). There is some evidence to suggest both pharmacological (varenicline, bupropion and nicotine replacement therapy) and behavioural interventions may be successful for supporting people with SMI to reduce smoking (Pearsall et al., 2018).
However, until Bekkering and colleagues’ (2026) systematic review and network meta-analysis, there was limited evidence comparing the effectiveness and safety of different smoking cessation interventions specifically in people with SMI.
Methods
Researchers searched several databases and trial registries up to September 2024 for randomised controlled trials (RCTs) enrolling adults who had established diagnoses of schizophrenia, bipolar disorder, recurrent or current major depressive disorder or post-traumatic stress disorder.
The trials also had to involve participants being randomised to a smoking cessation intervention versus another active treatment, placebo, standard care or no treatment.
All papers were screened by title, abstract and full text independently by two reviewers, and for eligible studies, one reviewer extracted data using a standardised, pre-tested form, with a second reviewer cross-checking for accuracy. A group of content experts and methodologists were blinded to the study results before grouping similar interventions into three arms:
- pharmacological interventions,
- non-pharmacological interventions, and
- combinations of pharmacological and non-pharmacological interventions.
The data was analysed using network meta-analysis, which allowed the researchers to compare the effectiveness and safety of different cessation treatments, including some treatments that had not been directly compared in individual trials. For the purpose of analysis, the researchers combined placebo and minimal care. The study team also used the Cochrane risk-of-bias tool version 1 to review risk of bias and the GRADE approach for assessing the certainty of the evidence.
Results
Overall, 74 studies of 11,023 participants were included in the review (mean age=46; 65% of participants were male). The trial populations included those with schizophrenia spectrum disorders (35 studies), post-traumatic stress disorder (PTSD) (14 studies), bipolar disorder (5 studies), or a mixed population of people with severe mental illness (21 studies). All trials were conducted in high-income countries. At baseline, participants smoked a mean of 20 cigarettes per day and across all studies, the average treatment lasted around 12.5 weeks, with follow-up extending for about 22 weeks after treatment.
27 of the RCTs reporting on short-term smoking abstinence were included in the network meta-analysis. Of these, three interventions were supported by moderate- or high-certainty evidence as being more effective than placebo/minimal care:
- Varenicline (OR= 3.89, 95% CI [2.65 to 5.81]); this translates to 17 more quitters per 100 adults.
- Nicotine replacement therapy (NRT) (OR= 2.30, 95% CI [1.54 to 3.54]); 8 more quitters per 100 adults.
- Bupropion (odds ratio [OR]= 2.01, 95% confidence interval [CI] [1.30 to 3.11]); 7 more quitters per 100 adults.
Varenicline was better than NRT (OR= 1.69, 95% CI [1.14 to 2.44]) and bupropion (OR= 1.92, 95% CI [1.32 to 2.93]). These estimates were found to be consistent across the sensitivity analyses and were supported by moderate or high certainty evidence.
For long-term smoking abstinence, 28 trials were included in the network meta-analysis. Two interventions, supported by moderate or high certainty evidence, were reported to be better than placebo/minimal care: bupropion (OR= 1.89, 95% CI [1.17 to 2.96]); and varenicline (OR= 2.84, 95% CI [1.86 to 4.27]). A combined pharmacological and non-pharmacological intervention was supported by low certainty evidence (OR= 2.09, 95% CI [1.49 to 3.00]).
Notably, among the trials reporting depressive symptoms and psychopathology, there was little or no difference with varenicline, bupropion or NRT compared with placebo or minimal care. This is reassuring, although it does not establish that these treatments carry no risk, as the evidence regarding serious adverse events remains very uncertain. For combined pharmacological and non-pharmacological treatments, the evidence was less certain, with possibly little or no difference compared with placebo or minimal care.
The researchers also looked at a few other factors, including serious adverse events. The meta-analysis of 11 trials suggested a small increase in risk for all interventions when compared with placebo/ minimal care, but the evidence was very uncertain across all treatments. For dropout to harm, quality of life and weight gain, analyses showed little to no difference across all comparisons, but again, the certainty of this evidence was low.
Conclusions
The results of the network meta-analysis indicate that when compared with placebo or minimal care, there was strong evidence for using varenicline for smoking cessation for people with SMI to increase both long- and short-term abstinence. This finding matters because people with SMI should have equal access to disease prevention and the opportunity to have the same physical health as others. Bupropion and NRT were also found to improve short-term abstinence, but the evidence for these was somewhat less certain.
Combined pharmacological and non-pharmacological treatments may improve the chances of achieving long-term smoking abstinence for people with SMI, although the certainty of this evidence was low. Kypriotakis et al. (2024) found a similar result in their secondary analysis of the EAGLES RCT, where varenicline plus behavioural counselling was highlighted as an effective cessation option for individuals living with major depressive disorder (you can find out more about this study in Gemma Taylor’s Mental Elf blog). Additionally, a Cochrane global review found behavioural support involving economic incentives or counselling consistently demonstrated benefits for smoking cessation (Hartmann-Boyce et al., 2021). Future research should focus on high-quality RCTs testing combined pharmacological and behavioural support for people with SMI, as the evidence in this review was uncertain.
Strengths and limitations
One of the main strengths of this review is the large, comprehensive data set of 74 studies with 11,023 participants. This means the analysis was well powered, making the statistical findings more reliable. Another strength is the use of a network meta-analysis, which allowed the researchers to compare across multiple different cessation strategies, even if there weren’t any trials directly comparing them. The network meta-analysis allowed comparison of interventions that had not necessarily been directly compared, while incorporating the certainty of the evidence. The broad search strategy and strict adherence to the methodology (i.e., duplicate screening, use of Cochrane risk-of-bias tool version 1, use of the GRADE approach, subgroup and sensitivity analyses, etc.) further increase the reliability of the review’s findings. Additionally, including additional patient outcomes like mental health, quality of life, weight gain and serious adverse events allows a more real-world, applicable overview of the different treatment options, which would be supportive to both patients and clinicians in decision-making.
However, there are a few limitations with the study:
- Over a third of participants in the review come from the EAGLES trials, which enrolled around 4,166 participants with psychiatric disorders. The main issue with this is that it limits the generalisability of the findings to other settings and populations and means the results are more dependent on the assumptions and methods used in the EAGLES trials.
- Another factor which limits the generalisability of the findings is that most studies included in the review were conducted in high-income countries. This is not something the researchers had control over, but should be considered when thinking about applicability.
- Out of the 74 studies, 72 were at risk of bias for at least one domain. The main issues were unclear reporting of randomisation procedures and lack of masking (participants, professionals and assessors). This makes it more likely that the treatment effects could be overestimated or underestimated.
- The review identified only one study on the effectiveness of e-cigarettes as a smoking cessation aid for people with SMI, so it is difficult to judge whether this is a viable option for this population. Encouragingly, the University of Oxford is testing a trial that combines a brief offer of an e-cigarette with behavioural support for people with severe mental illness (you can read more about the study here TRIDENT). The evidence from this trial should help inform future treatment options.
- Another limitation was that the non-pharmacological interventions varied widely in content, intensity, and setting, making it hard to identify which type of behavioural support works best for people with SMI. Future research should compare different formats and intensities more directly.
Implications for practice
The findings support discussing varenicline as an effective smoking cessation option with people living with SMI, particularly given the moderate to high-certainty evidence for long-term abstinence. Interestingly, current evidence suggests varenicline is less likely to be prescribed for people living with SMI than other medicines, like NRT (Taylor et al., 2020). Some clinicians may remain cautious about prescribing varenicline to people with SMI because of concerns about psychiatric adverse effects, as listed in NICE guidelines. The current review found there was little or no effect on depression or overall psychiatric symptoms with varenicline, bupropion and NRT compared with placebo or minimal care, which should provide some reassurance, although evidence regarding serious adverse events remains very uncertain. Future research should explore these issues further, and help inform future updates to clinical guidance and support the implementation of existing NICE recommendations.
Current guidelines also recommend using a combination of pharmacological and non-pharmacological interventions for smoking cessation for SMI (SAMHSA, 2019); however, the findings of the meta-analysis indicate uncertainty about the additional benefits of using both treatments. This uncertainty should not be interpreted as evidence that behavioural support is ineffective. Clinicians can continue to offer both pharmacological and behavioural options according to patients’ preferences, needs and availability of support, including pharmacological treatment for people who decline behavioural support. The meta-analysis evidence was also unclear for serious adverse events, quality of life and weight gain outcomes across the different smoking cessation interventions. Clinicians should be open and honest with patients about the uncertainty around these outcomes when discussing treatment options and use shared decision-making where appropriate.
With this lack of evidence in mind, and the fact that several of the trials did not report on these outcomes at all, future research should focus specifically on the additional patient outcomes associated with the different treatments. This would provide a clearer picture of the risks and benefits of each treatment and aid in shared decision-making discussions between patients and clinicians. Further research should also explore whether the findings of this meta-analysis are applicable across different settings and countries, as access to smoking cessation medication, behavioural support and mental health services may be different in low- to middle-income countries, and, therefore, there might be variation in what treatment approaches are most effective or feasible.
Overall, this review highlights that several smoking cessation treatments can improve abstinence in people with SMI, with the strongest evidence supporting varenicline for long-term abstinence. However, evidence regarding serious adverse events and several other patient-important outcomes remains uncertain. Varenicline appeared to have the largest treatment effect among the principal pharmacological options studied, particularly for long-term abstinence. Large, well-designed RCTs are still needed to clarify the effects of combined pharmacological and non-pharmacological treatments, as well as the impact of the different treatments on additional patient outcomes.
Statement of interests
Kayley was not involved in this study, does not know the study authors personally, and was not involved in peer review or editorial decisions relating to publication of this paper. However, Kayley is currently working on the TRIDENT study – Tobacco RIsk reDuction with E-cigarette Nicotine Replacement Therapy. This programme aims to test the delivery of an e-cigarette and behavioural support programme for people with severe mental illness who smoke and have declined help to quit. AI tools were used for editing.
Edited by
Dr Dafni Katsampa.
Links
Primary paper
Geertruida Bekkering, Johan Detraux, José López-López, Marc De Hert, Eden Evins, Anne-Catherine Vanhove, Hanne Cloetens, Abdullah Algarni, Thomas Vandendriessche, Arnav Agarwal, Thomas Agoritsas, Bert Aertgeerts, Nicolas Delvaux & Mieke Vermandere (2026) Smoking cessation for people with severe mental illness: systematic review and network meta-analysis (PDF). BMJ Medicine 2026 5(1) e002190.
Other references
BNF. Varenicline. BMJ Publishing Group Ltd and the Royal Pharmaceutical Society of Great Britain. Retrieved July 21st 2026 from https://bnf.nice.org.uk/drugs/varenicline/
Brose, L., West, R., McDermott, M., Fidler, J., Croghan, E., & McEwen, A. (2011). What makes for an effective stop-smoking service? Thorax, 66(10), 924–926.
Callaghan RC, Veldhuizen S, Jeysingh T. et al (2014) Patterns of tobacco-related mortality among individuals diagnosed with schizophrenia, bipolar disorder, or depression. Journal of Psychiatric Research 2014 48(1) 102–110.
Hartmann-Boyce J, Livingstone-Banks J, Ordóñez-Mena JM. et al (2021) Behavioural interventions for smoking cessation: an overview and network meta-analysis. Cochrane Database of Systematic Reviews 2021 (1) CD013229.
Kypriotakis, G., Cinciripini, P., Green, C., Lawrence, D., Anthenelli, R., Minnix, J., Beneventi, D., Morris, C., Karam-Hage, M., & Blalock, J. (2024). Effects of Varenicline, Bupropion, Nicotine Patch, and Placebo on Treating Smoking Among Persons with Current or Past Major Depressive Disorder: Secondary Analysis of a Double-Blind, Randomised, Placebo-Controlled Trial. The American Journal of Psychiatry.
McDermott, M., Marteau, T., Hollands, G., Hankins, M., & Aveyard, P. (2013). Change in anxiety following successful and unsuccessful attempts at smoking cessation: cohort study. British Journal of Psychiatry, 202(1), 62–67.
Momen NC, Plana-Ripoll O, Agerbo E. et al (2022) Mortality associated with mental disorders and comorbid general medical conditions. JAMA Psychiatry 2022 79(5) 444–453.
Pearsall R, Smith DJ, Geddes JR. (2019) Pharmacological and behavioural interventions to promote smoking cessation in adults with schizophrenia and bipolar disorders: a systematic review and meta-analysis of randomised trials. BMJ Open 2019 9 e027389.
Public Health England. (2020). Health matters: smoking and mental health. Gov.uk website, last accessed 30 Jul 2026.
Royal College of Physicians. (2013). Smoking and mental health. rcp.ac.uk/media/2yudsitb/smoking_and_mental_health_-_full_report_web_0_0.pdf
Substance Abuse and Mental Health Services Administration (2019) Implementing tobacco cessation treatment for individuals with serious mental illness: a quick guide for program directors and clinicians (PDF). Rockville, MD: Substance Abuse and Mental Health Services Administration. SAMHSA Publication No. PEP19-02-00-001.
Taylor GMJ, Itani T, Thomas KH. et al (2020) Prescribing prevalence, effectiveness, and mental health safety of smoking cessation medicines in patients with mental disorders. Nicotine & Tobacco Research 2020 22(1) 48–57.
Taylor, G., McNeill, A., Girling, A., Farley, A., Lindson-Hawley, N., & Aveyard, P. (2014). Change in mental health after smoking cessation: systematic review and meta-analysis. BMJ, 348, g1151.