Statistics from Altmetric.com
Q Do antidepressants increase long term abstinence from smoking?
Clinical impact rating GP/FP/Primary care ★★★★★★☆ IM ★★★★★★☆ Respirology ★★★★★★☆
drug names found in the Tobacco Addiction Group’s specialised register, reference lists, and meeting abstracts were searched in PubMed and EMBASE/Excerpta Medica in 2002. Investigators were contacted.
Study selection and assessment:
randomised controlled trials (RCTs) that compared any antidepressant with placebo or another treatment and assessed abstinence of smoking at ⩾6 months. Studies were pooled using fixed effects.
30 RCTs were included. Tricyclic antidepressants. Nortriptyline increased smoking cessation (5 RCTs) (table). Monoamine oxidase inhibitors. 1 RCT of moclobemide did not show a statistically significant difference in abstinence at 12 months (table). Atypical antidepressants. 16 RCTs showed that bupropion was better than placebo for increasing abstinence (table). 2 RCTs compared bupropion with nicotine replacement therapy. 1 showed a benefit for bupropion, whereas the other unpublished RCT did not (table). 1 RCT compared bupropion with nortriptyline and showed no statistical difference (table). 1 study comparing venlafaxine with placebo did not show a statistically significant difference (table. Selective serotonin reuptake inhibitors (SSRIs). 4 RCTs (2 fluoxetine, 1 paroxetine, 1 sertraline) showed that SSRIs did not increase abstinence at ⩾6 months (table).
In smokers, bupropion and nortriptyline increase abstinence at ⩾6 months. Selective serotonin reuptake inhibitors do not increase abstinence.
Hughes et al live up to the Cochrane Collaboration’s reputation in this review, which shows benefit for 2 antidepressants (bupropion and nortriptyline) for long term abstinence in smoking cessation. This supports the US Public Health Service Guideline “Treating Tobacco Use and Dependence,” which gives a clear recommendation that all patients attempting to quit should be encouraged to use effective pharmacotherapy.1
While it is encouraging to learn of effective interventions, one aspect of this type of research may go unnoticed, which is the level of behavioural support that occurs. Typically, both experimental and control subjects receive counselling, but how much are they talking about? For example, in one high quality study included in the meta-analysis,2 all patients received individual counselling weekly for the first 7 weeks and at weeks 9 and 12, with telephone counselling at the fourth and fifth months. This could mean as much as 2 hours of individual counselling! Compare this with the level of behavioural support given in a busy clinical practice, which could be, “Here is your prescription, see you in 6 months, good luck.” If one wants to replicate the results of research in clinical practice, then one should try to replicate the methods under which the original research was conducted. Many factors may account for relapse, but when patients claim that “these medications don’t work,” physicians should inquire whether they were given with the appropriate amount of behavioural support.
Clearly, all smokers should be offered effective pharmacological assistance with quitting. This currently means 2 antidepressants (bupropion and nortriptyline) along with nicotine replacement therapy, for which there is overwhelming evidence of effectiveness.1 However, physicians also need to remember to include the appropriate amount of psychosocial and behavioural support, which gives our patients the best chance to achieve long term abstinence.
For correspondence: Professor J Hughes, University of Vermont, Burlington, VT, USA.
Sources of funding: National Institute on Drug Abuse (NIDA) and NHS Research and Development National Cancer Programme.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.