Smoking cessation interventions for smokers with current or past depression

  • Protocol
  • Intervention



This is the protocol for a review and there is no abstract. The objectives are as follows:

To evaluate the effectiveness of smoking cessation interventions for smokers with current or past depression.


In recent years, it has become increasingly clear that smokers are not a homogeneous group and that it is may be necessary to develop specific interventions for smoking cessation for specific groups (Pomerleau 1993). Smokers who have current or past depression have poor quit rates and may especially need help. The positive relationship between depression and smoking has been known for some time. In the 1970s it was shown that smokers are more neurotic than non-smokers and that smokers who had a successful quit attempt were less neurotic, depressed and anxious than those who did not succeed in their quit attempt (Eysenck 1973; Pomerleau 1979).

More recent studies have shown that compared to non-smokers, smokers often have current or past depression (Kalman 2005). This relationship between depression and smoking has been shown to be present as early as in adolescence (Brown 1996). Furthermore, although some research has shown that attempts to quit smoking fail considerably more often in smokers with current or past depression than in other smokers, whether this is true is debatable (Covey 2004; Hall 2004; Hitsman 2003; Hitsman 2004). Finally, when smokers with a past depression quit, they may have a significantly increased risk of developing a new depressive disorder compared to smokers without past depression (Glassman 2001).

Depression seems related to psychosocial characteristics that might make it even more difficult for those with depression to stop smoking, compared with those from the general population of smokers. For example, they may demonstrate a lower self-efficacy for quitting, be more likely to be unemployed, have a poorer network of social support and poorer physical health

The assumption that smoking and smoking cessation relapse are linked to a high incidence of current and past depression has resulted in several studies examining smoking cessation interventions that provide strategies for managing depressive symptoms (Brown 2001; Hall 1994; Hall 1996; Patten 1998). These smoking cessation interventions consist of pharmacological and/or psychosocial interventions. Among the latter, cognitive behavioural interventions in particular have been examined.


To evaluate the effectiveness of smoking cessation interventions for smokers with current or past depression.


Criteria for considering studies for this review

Types of studies

Randomized controlled trials (RCTs) and quasi-randomized trials.

Types of participants

Adult smokers with current or past depression, defined as having a current or historical clinical diagnosis of depression, according to the DSM-IV criteria.
We will also includes adult smokers who score on validated self-reported measures of depression (e.g. Beck Depression Inventory [BDI] or Center for Epidemiologic Studies Depression Scale [CES-D]).

Types of interventions

Any pharmacological or psychosocial intervention, alone or in combination.

Types of outcome measures

The primary outcome will be smoking status at a minimum of six months from the quit day. We will use sustained cessation rates in preference to point prevalence. Sustained cessation refers either to complete abstinence (continuous) or to prolonged abstinence (complete abstinence following a post-quit two-week period of grace in which minor relapse may occur) (Hughes 2003). We will use biochemically verified cessation in preference to self-reported quitting. The strictest definition of smoking cessation reported will be used in meta-analysis. We may examine types of abstinence (continuous, prolonged, point prevalence) and levels of verification (biochemical, self-report) as potential modifiers of outcome. We will regard smokers lost to follow up as being continuing smokers, where such data are available.

The secondary outcome will be change in depression status according to diagnostic criteria.

Search methods for identification of studies

See the search strategies for the Cochrane Tobacco Addiction Group and the Cochrane Depression, Anxiety and Neurosis Group.

All relevant trials meeting our inclusion criteria will be identified by:

  • A computer-aided search of MEDLINE (from 1966 to April 2006), EMBASE (from 1989 to April 2006) and PsycINFO (from 1971 to April 2006) databases, using the search strategies recommended by the Cochrane Tobacco Addiction Group and the Cochrane Depression, Anxiety and Neurosis Group;

  • Screening references given in relevant reviews and identified RCTs (e.g. citation tracking);

  • Screening of the Cochrane Central Register of Controlled Trials (CENTRAL), Issue 2, 2006 (or the most recent Issue);

  • Screening of personal bibliographies and communication with experts in the field.

We will consider unpublished studies or studies found only as abstracts where sufficient detail is available. We will contact trial authors for further data if necessary.

Data collection and analysis

Study selection
In order to select the studies published up to April 2006 RvdM and MW will run a comprehensive search. The same authors will independently select the studies to be included in the systematic review, by applying the agreed selection criteria. We will resolve disagreements concerning selection and inclusion of RCTs by consensus where possible. We will consult with FS and PC if disagreements persist.

Methodological quality assessment
RvdM and MW will use the the Delphi List (Verhagen 1998) to assess the methodological quality of selected studies, independently assessing the methodological quality of the studies, using full text versions, except for abstracts to be included. The articles will not be blinded for authors, institution and journal, because the review authors who will perform the quality assessment are familiar with the literature. If an article does not contain enough information on methodological criteria, i.e., if one or more criteria are scored 'unclear', we will contact the trial authors for additional information.

Data extraction and analysis
RvdM and MW will independently extract the data using a standardized form. We will pilot the data extraction methods using two RCTs on smoking cessation. If data allow, we will conduct a meta-analysis of the included studies, using the Mantel-Haenszel Odds Ratio, fixed-effect method. We will use the I2 statistic (Higgens 2003) for testing the null hypothesis of homogeneity across the primary studies in the meta-analysis. Similarly, under a random effects model we will use the percentage of the variance accountable by sample error as an index for homogeneity.

To explore reasons for the observed heterogeneity, and if the data allow, we will perform sensitivity analyses on study characteristics (e.g. type of randomization, blinding, losses to follow up, measurement instruments), population characteristics (e.g. percentage female, mean age, diagnosis), publication characteristics (e.g. publication year, publication type), intervention characteristics (e.g. types of treatment, reference treatment, number of sessions), outcome characteristics (e.g. outcome domain, reliability of the measurement instruments, follow-up time), and test their influence on the association between the smoking cessation intervention and cessation of smoking.
We will calculate odds ratios, relative risks, risk differences and 95% confidence intervals for every study, where the data are available.


We would like to thank John Hughes and Michael Ussher for reading and commenting on earlier drafts of this protocol.

What's new

21 October 2008AmendedConverted to new review format.


Protocol first published: Issue 3, 2006

Contributions of authors

RM van der Meer and MC Willemsen will identify and select all the studies. Both reviewers will also assess the methodological quality of the studies and perform the data extraction. RM van der Meer, MC Willemsen and F Smit will conduct the data analysis. F Smit and P Cuijpers will be consulted in case of disagreements with regard to the quality assessment and/or data extraction. They will also be involved in the final decisions regarding in- and exclusion of studies, and with regard to judgements about results and conclusions. All authors will be involved in writing the review protocol and the final review.

Declarations of interest

None known

Sources of support

Internal sources

  • STIVORO - for a smokefree future, Netherlands.

  • Vrije Universiteit Amsterdam, Netherlands.

  • Trimbos-Instituut, Netherlands Institute of Mental Health and Addiction, Netherlands.

External sources

  • No sources of support supplied