A rapid review of the effectiveness of smoking cessation interventions for people with anxiety and/or depression living within the community

1.1 Who is this review for?

This Rapid Review was conducted as part of the Health and Care Research Wales Evidence Centre Work Programme. The research question was formulated by Welsh Government Health Improvement to inform the Welsh Government delivery plan for 2024-2026.

1.2 Background and purpose of this review

Smoking remains a significant public health concern in Wales. It is the leading cause of preventable death in Wales and costs the Welsh NHS an estimated £302 million annually (Welsh Government 2022). The Welsh Government has published a long-term plan towards a smoke-free Wales, with the aim of reducing smoking prevalence from the current rate of 13%, to below 5% of the population by 2030 (Welsh Government 2022).

People with mental health conditions, including those with anxiety and depression, have significantly higher smoking rates compared to the general population; in Wales, 35% of adults aged 16 and over who report having a long-standing mental health condition are smokers (Welsh Government, 2023a). This is over double the rate of smokers in the general population of Wales (13%) (Welsh Government 2023b). These populations are also less likely to access smoking cessation services (National Institute for Health and Care Excellence, 2021). Furthermore, the life expectancy of people with poor mental health is on average 10 to 20 years less than the general population, with smoking being the biggest cause of this difference (Public Health England, 2020).

Anxiety and depression together are the most prevalent common mental health disorders (NHS Digital, 2016). According to the most recent Adult Psychiatric Morbidity Survey, 5.9% and 3.3% of adults aged 16 years and over meet the diagnostic criteria for anxiety or depression, respectively (NHS Digital 2016). Additionally, recent data collected by the Office for National Statistics (2021), showed 16% of adults in Great Britain reported they were experiencing depressive symptoms, and 16% were also likely to be experiencing some form of anxiety.

A range of smoking cessation interventions have been shown to be effective in the general population. According to two Cochrane systematic reviews, these include behavioural interventions such as counselling and guaranteed financial incentives (Hartmann-Boyce et al, 2021), and pharmacological interventions like Varenicline, Bupropion, cytosine, nicotine replacement therapy (NRT), and nicotine e-cigarettes (Lindson et al, 2023). It is unclear if these interventions would also be effective in people with anxiety and depression. Wales’ national smoking cessation service, Help Me Quit, currently offers six weeks of behavioural support and 12 weeks of NRT to people who smoke however its effectiveness among people with anxiety and depression is largely unknown. Research suggests that people with mental health disorders are likely to have a better quality of life and feel more positive and calmer after quitting smoking (NHS, 2024). The beneficial effects of giving up smoking on anxiety and depression symptoms is said to equal that of taking anti-depressant medication (NHS, 2024). There is therefore a need to identify the most effective smoking cessation interventions for this population group.

This rapid review focuses on primary studies and aims to identify and synthesise the evidence for the effectiveness of smoking cessation interventions for people with anxiety and/or depression living in the community.

2.1 Overview of the Evidence Base

Eleven studies (15 reports) were eligible for inclusion in this rapid review. Eligibility was restricted to comparative primary studies assessing the effectiveness of smoking cessation interventions for adult smokers with anxiety or depression living in the community (see section 5.1 for full eligibility criteria). Ten of the 11 studies were randomised controlled trials (RCTs), two of which were pilot studies. One study was a quasi-experimental uncontrolled before and after pilot study. Studies were conducted in a range of countries, including USA (n=6), Spain (n=1), France (n=1), Netherlands (n=1), and two studies were conducted across multiple countries across the EU and USA. Sample sizes varied across the 15 identified reports, ranging from 17 to 6,653, and 11 reports included more than 100 participants. Reports were published between 2008 and 2023, nine since 2019.

A protocol for a UK-based study investigating the acceptability and impact of a smoking cessation intervention for mental health service users with depression, alongside routine psychological therapy was also identified during searching. This study may be particularly useful for informing smoking cessation interventions for people with anxiety and/or depression as it is UK based and is incorporated into existing psychological therapy for mental health service users. We were informed by the corresponding author that the study report is currently under consideration with the Addiction Journal. The protocol is registered on the International Standard Randomised Controlled Trial Number (ISRCTN); ISRCTN99531779 (https://doi.org/10.1186/ISRCTN99531779).

All of the studies included participants who had depression (n=11), one of which also included people with anxiety. A range of diagnostic tools were used to identify study participants with depression and anxiety at study entry. This included the Structured Clinical Interview for axis I disorders of the DSM-IV, the Beck Depression Inventory-II (BDI-II), the neuropsychiatric interview, the Centre for Epidemiological Studies Depression (CES-D) scale, the Patient Health Questionnaire (PHQ) versions 8/9, and the Hospital Anxiety and Depression Scale (HADS).

For the synthesis of results, studies have been grouped into five categories according to the type of intervention(s) being assessed: psychological (n=3), pharmacological (n=3), exercise-based (n=2), multicomponent pharmacological and psychological (n=2), and multicomponent exercise and psychological interventions (n=1). Greater detail of the specific interventions being investigated is shown in Table 1, which also outlines the current treatment used in Wales. All intervention components only received by participants in the experimental group are highlighted in green (an Asterix is also given for readers with difficulty distinguishing colours). In most studies (n=8), the interventions being assessed were an addition to existing smoking cessation treatment which was also received by participants in the control group. The smoking cessation treatment components received by all participants (intervention and control groups) are highlighted in yellow in Table 1 (or X without an Asterisk).

Interventions were delivered at clinical trials centres or outpatient clinics (n=3), research clinics (n=2), a hospital (n=1), and remotely via a mobile phone application, text messaging programme, or telephone (n=2). Three studies did not state the intervention setting.

Intervention durations ranged from six weeks to three months. Details of the interventions can be seen in Table 2; a summary of the findings can be found in Table 3, and a more detailed summary of the included studies can be seen in section 6.2 (Table 5).

Outcome measures identified included smoking abstinence-related outcomes; 7-day point prevalence abstinence (PPA), quit rates, continuous or prolonged abstinence, smoking reduction outcomes; quit attempts, number of cigarettes smoked per day, and cigarette demand, anxiety or depressive symptoms, adverse events, and cost-effectiveness. Data collection methods included self-reported smoking and abstinence rates, validated questionnaires, and tests for biochemical verification were performed in all studies except two (Dahne et al, 2012; van der Meer et al, 2010).

The methodological quality of included studies was assessed using the appropriate Joanna Briggs Institute (JBI) critical appraisal tool for RCTs or quasi-experimental studies. All reports were judged to have some risk of bias, often because of poor reporting of methods. Of the 14 RCT reports, 11 were considered low quality, one was considered moderate quality, and two were considered high quality. Concerns which increased the risk of bias of RCTs included uncertainty about measurement reliability in 11 reports, no description or analysis of loss to follow-up in 11 reports, absence of power calculations stated in 11 reports, and no details about blinding of assessors in 10 reports. The quasi-experimental uncontrolled before and after pilot study was considered low quality due to the absence of a control group and power calculation, therefore it was unclear whether it was sufficiently powered to measure intervention effectiveness. Further details of the quality appraisal can be found in section 6.3.

Multiple reports presented findings from three studies (Anthenelli et al, 2013; Anthenelli et al, 2016; Weidberg et al, 2018), which were included because they provided additional, unique results relevant to the research question. Reports of the same study were identified by trial registration numbers and grouped accordingly. Doran et al (2019) conducted a secondary analysis of the international study reported by Anthenelli et al (2013) (trial number: NCT01078298). Both Ayers et al (2019) and Cinciripini et al (2022) were secondary analyses of the international Evaluating Adverse Events in a Global Smoking Cessation Study (EAGLES) (NCT01456936) (Anthenelli et al, 2016). However, Ayers et al (2019) compares the findings for participants with and without anxiety disorders, whereas Cinciripini et al (2022), compared the findings for participants with and without depression. Anthenelli et al (2016) was not eligible for inclusion because smoking cessation outcomes were not reported specifically for participants with either anxiety or depression. González-Roz et al (2021), Weidberg et al (2018), and Secades-Villa et al (2019), all report findings from the same Spanish study (NCT03163056), although Weidberg et al (2018) report findings from a smaller sample of participants. For the synthesis and summary of study findings (Table 3), reports of the same study have been grouped together.

2.2 Impact of psychological interventions

Five reports of three studies (González-Roz et al, 2021; Heffner et al, 2019; Secades-Villa et al, 2019; van der Meer et al, 2010; Weidberg et al, 2018) were examining the effect of psychological interventions for improving smoking cessation and mental health outcomes in smokers with depression. Three of the five reports (González-Roz et al, 2021; Secades-Villa et al, 2019; Weidberg et al, 2018) were analyses of the same smoking cessation study conducted in Spain. The remaining studies were conducted in the Netherlands (van der Meer et al, 2010), and USA (Heffner et al, 2019). All studies were RCTs except for one quasi-experimental before and after pilot study (Heffner et al, 2019). One RCT was considered high quality (van Der Meer et al, 2010), the remaining studies were considered low quality. All studies focused on participants with depression. The findings of studies evaluating the effectiveness of psychological smoking cessation interventions was mixed.

2.3 Impact of pharmacological interventions

Five reports were examining three different studies exploring the effect of pharmacological interventions on smoking cessation and mental health outcomes (Anthenelli et al, 2013; Ayers et al, 2019; Cinciripini et al, 2022; Doran et al, 2019; Evins et al, 2008). Four reports were of two studies two studies conducted internationally across the EU and USA (Anthenelli et al, 2013; Anthenelli et al, 2016), and one study was conducted in the USA alone (Evins et al, 2008). One study was considered high quality (Anthenelli et al, 2013), while the remaining studies were considered low quality. Four reports focused on participants with depression, and one on participants with anxiety (Ayers et al, 2019). Findings in relation to the effectiveness of the interventions appear mixed.

2.4 Impact of aerobic exercise-based interventions

Two studies assessed the impact of aerobic exercises of varying intensities on smoking cessation and mental health (Abrantes et al, 2023; Patten et al, 2017). Both studies were RCTs conducted in the USA and included participants with depression. One study was considered high quality (Abrantes et al, 2023), while the other was considered low quality. Findings in relation to the impact of the interventions appear mixed.

2.5 Impact of multicomponent interventions comprising of pharmacological and psychological components

Two studies investigated the effect of a combination of pharmacological and psychological interventions on smoking cessation and mental health outcomes (Dahne et al, 2023; Hitsman et al, 2023). Both studies were RCTs conducted in the USA and recruited participants with depression. Findings appear to be in support of the interventions.

2.6 Impact of multicomponent interventions comprising of exercise and psychological intervention components

One study examined the use of a multicomponent intervention comprising of an exercise and behavioural component (Bernard et al, 2015). This study was a pilot RCT conducted in France and focused on smokers with depression. It was considered low quality, and the findings did not favour the intervention.

3.1 Summary of the findings

This rapid review aimed to identify and synthesise the evidence for the effectiveness of smoking cessation interventions in people with anxiety or depression living in the community. There is some, mostly low quality, evidence to suggest that smoking cessation interventions may be effective at improving a range of smoking cessation and mental health outcomes in people with depression. Only one study was identified assessing smoking cessation interventions in people with anxiety meaning conclusions about the evidence of effectiveness cannot be drawn. Psychological interventions may increase abstinence in smokers with depression, but results were mixed regarding their impact on mental health outcomes. Pharmacological interventions, particularly Varenicline, appear effective for smoking cessation in people with depression and anxiety. However, evidence for Bupropion and NRT is limited and mixed. Findings generally did not support the effectiveness of exercise interventions for smoking cessation, and no sustained benefits or improvements in depressive symptoms were identified. Although the evidence was limited, none of the studies suggested that the interventions caused greater harm than the smoking cessation treatment received by the control group. Delivering interventions as part of a multicomponent approach may be beneficial, but methodological limitations and variation in intervention types and outcome measures across studies compromise the applicability of these findings. Further research with rigorous methodologies is needed to provide clearer guidance on effective smoking cessation strategies for this population.

3.2 Strengths and limitations of the available evidence

A strength of the available evidence is the identification of a wide range of smoking cessation interventions. However, the considerable heterogeneity across studies makes it difficult to compare interventions directly and limits the overall certainty of the evidence base.

Variation in recruitment strategies raises questions about the similarities of participants across the included studies. Furthermore, inconsistency in diagnostic tools and cut-offs scores for inclusion hinders comparisons of depression severity and intervention effectiveness across studies.

The rapid review identified limited, low-quality evidence for smoking cessation interventions in individuals with anxiety or depression. Most studies lacked methodological rigor and were poorly reported. Only two RCTs were deemed high quality, while industry funding and involvement in five studies raised concerns about potential bias. Although 11 studies were identified, three were small pilot studies. These pilot studies, while potentially valuable for informing future research directions, were not powered sufficiently to detect intervention effectiveness due to their limited sample sizes. Additionally, seven reports were secondary analyses of three primary studies. Although secondary analyses may provide data from original studies with greater relevance to the research question (e.g., depression sub-group analysis), they may not be sufficiently powered and there is potential for overlap of participants across reports. If multiple reports draw data from the same original study population, this can inflate the overall sample size but not necessarily represent new data or a broader participant pool. Furthermore, data is limited by the original studies’ research questions and data collection methods. These limitations might mean the data wasn’t specifically collected to answer questions related to smoking cessation interventions for individuals with anxiety or depression.

Several studies included in this rapid review involved interventions being assessed as an addition to existing smoking cessation treatment which was received by all participants. It is important to consider that the relative effect of the intervention compared to the control may be diminished due to the underlying smoking cessation treatment received by all participants. While examining the effect of an intervention as an addition to existing treatment allows for an assessment of specific intervention components, the underlying treatment received by all participants likely affects the overall findings.

Several evidence gaps were identified. Notably, none of the included studies were conducted within the UK. This absence of recent UK-based research limits our understanding of how these interventions might translate to the specific healthcare system and cultural context of the Wales. Although no published UK studies met our criteria, a protocol for a UK-based smoking cessation intervention for smokers with depression was identified. This study, currently under review for publication in the Addiction journal (ISRCTN99531779; https://doi.org/10.1186/ISRCTN99531779), is especially pertinent to this rapid review as it explores smoking cessation support integrated into routine psychological care within the UK. Additionally, the review identified gaps relating to barriers and facilitators to smoking cessation interventions, and the populations studied. Only one study focused on participants with anxiety disorders. No studies reported co-existing mental health conditions, which would have provided a more nuanced understanding of intervention effectiveness within this population. These limitations suggest a need for caution when generalising the findings to broader populations.

This rapid review found no evidence on equity factors and intersectionality. Included studies did not consider how socioeconomic factors, ethnicity, or social capital affect access and success of interventions. Understanding these complexities is essential for developing inclusive and equitable smoking cessation programs.

Additionally, the rapid review found no studies investigating harm reduction approaches, and limited research reporting interventions applied at critical touchpoints within healthcare systems and smoking cessation services (where targeted interventions could reach the review’s included population), or interventions that integrate with existing healthcare pathways or mental health services.

This rapid review identified no studies evaluating the effectiveness and safety of e-cigarettes as a smoking cessation or harm reduction tool for people with anxiety or depression. This highlights an absence of research on e-cigarettes as a smoking cessation or harm reduction aid in our population of interest. E-cigarettes have become a popular alternative to traditional cigarettes, potentially offering a route to harm reduction or smoking cessation for smokers with depression or anxiety. Despite their potential benefits, the rising prevalence of e-cigarette use among non-smokers, particularly children and young people, raises concerns which requires further investigation.

Only one study included in this rapid review analysed intervention cost-effectiveness, which investigated the one-year cost-effectiveness of adding contingency management to cognitive-behavioural treatment and behavioural activation for quitting smoking in smokers with depression in Spain. Despite reporting positive outcomes however, we cannot comprehensively assess the cost-effectiveness of such interventions due to insufficient evidence.

3.3 Strengths and limitations of this Rapid Review

The studies included in this rapid review were systematically identified through an extensive search of electronic databases and grey literature. The review was also based on an abbreviated systematic review approach, following established guidelines for conducting rapid reviews in an attempt to capture all relevant publications with minimal risk of bias in a timely manner.

Due to the short timeframe to complete this rapid review, as well as the large number of references initially identified from literature searches, most reports were screened by a single reviewer at title and abstract level. However, 558 references (14%) were double screened to assess consistency. The remaining stages of the review process included consistency checking by at least one other reviewer, with issues being discussed and resolved within the team.

A key strength of this rapid review is its focus on using rigorous participant selection criteria for identifying relevant studies. Studies were only considered eligible if they recruited participants using validated measurement tools for depression and anxiety, and specific cut-off points, depending on the tool used. This approach enabled us to include people with a clinical diagnosis already within the healthcare system, and those who should be but are not currently in the system. Although these two groups may have different needs and respond differently to smoking cessation interventions, we were able to capture a more comprehensive picture of the effectiveness of these interventions within a population accessing the healthcare system.

However, limitations exist. It is important to note that many studies excluded participants who were currently prescribed medication for depression or anxiety. This exclusion may have inadvertently left out individuals who are currently within the healthcare system and who may have responded differently to the intervention than those included in the study.

We excluded primary studies that investigated intervention effectiveness by comparing those with anxiety or depression with a non-psychiatric group (see Appendix 2). This was because our rapid review focussed on people with anxiety and depression, therefore we wanted to compare effectiveness in this population only. However, including these studies may have provided a more detailed understanding of intervention effectiveness.

Furthermore, the review excluded studies relying solely on self-reported depression, potentially missing individuals who experience symptoms but lack a formal diagnosis – a potential hard-to-reach group likely to have been of interest to our stakeholders. Second, studies focusing on anxiety sensitivity were also excluded. While anxiety sensitivity is often related to anxiety disorders, we excluded studies focused on this population to maintain focus on interventions specifically targeting people with anxiety or depression, ensuring our findings provide clear conclusions relevant to these populations. Finally, the rapid review did not include studies investigating mixed mental health populations unless they provided separate analyses for participants meeting specific criteria for anxiety or depression. These limitations restrict the generalisability of the findings to a broader population struggling with co-occurring mental health challenges and potentially experiencing symptoms that fall outside the strict diagnostic categories employed in the rapid review.

3.4 Implications for policy and practice

This rapid review identified a lack of high-quality evidence assessing the effectiveness of smoking cessation interventions for individuals with anxiety or depression. However, despite this, findings identified could help to inform future intervention development and delivery in this underserved population. It is, however, important to consider the context in which the interventions outlined here were evaluated and the uncertainty of the findings as outlined in table 3.

The rapid review identified Varenicline as a potentially effective smoking cessation medication. Varenicline was manufactured by Pfizer, who withdrew it from the market in 2021. More recently, another manufacturer applied for approval from the Medicines and Healthcare products Regulatory Agency (MHRA) for a generic Varenicline. Additionally, as of 09/07/2024, a different treatment with similar action to Varenicline called Cytosine, has been recommended for use in the general population by the All Wales Medicine Strategy Group, and is expected to be available on prescription within two months.

3.5 Implications for future research

Our rapid review identified a paucity of UK-based research in this area, indicating a need for future investigations within the UK to comprehensively assess the effectiveness and generalisability of interventions, particularly within the Welsh context.

While several RCTs were identified, the risk of bias assessment revealed considerable methodological limitations. Future research should prioritise conducting high-quality research with robust methodologies to minimise bias and improve the overall certainty of the evidence base.

There is a notable gap in research on interventions for individuals with predominant anxiety disorders. Future studies should focus on tailored interventions for this population group.

Additionally, our review excluded those self-reporting depression or anxiety in order to identify a population most similar to those engaged with mental health services. Studies which included populations with mixed mental health conditions were also excluded from our review if no sub-group analyses was conducted specifically in participants with either anxiety and/or depression. However, most studies with mixed mental health populations did not perform such analyses. Therefore, future research that includes participants with a range of mental health conditions should conduct analyses to identify the effectiveness of interventions for participants with specific mental health conditions, as it is likely that their needs for support to aid quitting smoking, and the effectiveness of interventions, will vary. This will help to provide greater insight to inform policy for people with specific mental health conditions.

While the rapid review identified various interventions for smoking cessation in individuals with depression, translating efficacy trials into real-world practice remains challenging. Prioritising effectiveness trials within real-world settings is crucial for developing impactful interventions tailored to the Welsh context. In particular, policymakers need evidence-based approaches that can be incorporated into the current system in Wales, enabling them to leverage existing resources.

To strengthen research on smoking cessation interventions for individuals with depression or anxiety, standardised measurement using validated diagnostic tools is essential. Transparent reporting of eligibility criteria will facilitate comparisons to real world settings and build a more robust evidence base. Future research should also include full economic evaluations or cost-effectiveness analyses to provide insights into the financial implications of implementing these interventions within healthcare systems.

3.6 Economic considerations*

• The total economic cost of smoking in people with mental health disorders in the UK is estimated at £3.5 billion** per annum. Of this total cost, £1.1 billion** is attributed to smoking-related healthcare treatment, £1.2 billion** to absenteeism and productivity losses and £1.2 billion** to premature mortality (Wu et al., 2015).

• There is limited economic evidence on the impact of implementing smoking cessation interventions for individuals living with depression and/or anxiety.

• Economic modelling evidence suggests tailored smoking cessation interventions combining behavioural and pharmaceutical support for individuals with severe mental illness can be cost-effective compared to usual care. Severe mental illness in this study concerned any mental illness causing significant functional impairment and limits on major life activities (Mattock, Owen & Taylor, 2023).

*This section has been completed by the Centre for Health Economics & Medicines Evaluation (CHEME), Bangor University

** Figures have been inflated to May 2024 prices using the Bank of England Inflation Calculator. These figures assume no difference in observed patterns since the original calculation.