How to Quit Smoking

It is difficult to quit smoking.

Here is my summary of the evidence about things that help:

The single most effective treatment to help smokers quit is a new drug called varenicline. This drug works by mildly stimulating a nicotine receptor, while blocking nicotine itself from interacting with the receptor: in this way it is a "nicotine receptor partial agonist." Varenicline is quite well-tolerated, the most common side-effects being nausea and insomnia. Usually these settle with time, and are less a problem if the dose is started low, and built up gradually. There have been reports of adverse psychiatric side effects (e.g. agitation, worsened insomnia, worsened depression) so it would have to be used cautiously in those with mental illnesses. I have reviewed a few studies below which affirm its usefulness among patients with psychiatric problems.

Evidence shows that there is only about a 10% chance of being able to quit smoking on your own (by quitting, we mean staying abstinent for at least a year).


A 3-month course of bupropion (an antidepressant) approximately doubles your chance of being able to quit. However, this raises your chance only to about 20%.

Tricyclic antidepressants such as nortriptyline can increase abstinence rates, probably comparable to bupropion.

Nicotine replacement (e.g. gum or patch) is less effective than bupropion. But it does increase your chances of quitting to about 15%.

Varenicline is most effective of all; a 3-month course increases your chance of quitting to about 25%.

Probably, combinations of the above pharmacological treatments increase your chances further.

Also I should note that many of the studies looking at pharmacological treatments for smoking addiction only used the active treatment for three months. It seems to me that longer courses of treatment would be more likely to help people maintain sustained abstinence; addictions and other long-standing phenomena in the brain persist, or change, over a course of years, not just months.

Psychotherapeutic strategies (e.g. CBT and other behavioural therapies) may help, but the evidence is weaker. The evidence that is available suggests that if psychotherapeutic or motivational strategies are to be effective, they need to be maintained over the long-term (perhaps permanently). In this regard, it reminds me of a "12-step" philosophy, which emphasizes the permanence of an addictive problem, and emphasizes that lifelong vigilance is needed to prevent relapse.

The following study published in CMAJ showed 54% 1-year abstinence in a group of smokers who had suffered an heart attack (MI), and who were given an "intensive anti-smoking intervention" (advice, an hour of counseling, and 7 telephone follow-up sessions over 60 days). The counseling employed "Marlatt and Gordon's relapse prevention model." A similar group of smokers not receiving this intervention had a 35% 1-year abstinence rate. Interestingly, medications were permitted in this study, and were associated with markedly worse abstinence rates. But the medications were administered more or less ad lib, so the effect of medications would be very confounded and unclear (for example, perhaps only the patients struggling most would have opted for medications--the reason they didn't do as well is because they were more severely addicted in the first place, not because of the medications. Also, with a haphazard administration of medications, patients might not realize the need to continue medications longer-term to maintain a therapeutic effect).
http://www.ncbi.nlm.nih.gov/pubmed/19546455
In my opinion, the level of "intervention" here actually seems quite minimal, yet it seems impressive that an organized effort of any kind to help prevent smoking through counseling methods would produce good results.

This is the best review article about medication treatments to date, in my opinion; it is from The Canadian Medical Association Journal (July 2008):
http://www.ncbi.nlm.nih.gov/pubmed/18625984?ordinalpos=87&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DefaultReportPanel.Pubmed_RVDocSum
A current study by Michael Steinberg et al. in Annals of Internal Medicine (2009;150:447-454) shows that combination therapy with bupropion + nicotine patch + nicotine inhaler, increased abstinence rates at 26 weeks to 35% in a group of medically ill smokers, compared to 19% in a group receiving only a nicotine patch. Those in the combination group were encouraged to use the treatments as long as they felt necessary, then to taper and discontinue as they felt able. This instruction, in my opinion, would have discouraged the participants from considering that bupropion could work to prevent relapse in the long-term, therefore they would probably have chosen to discontinue the bupropion as soon as they felt free of their smoking habit for a short time. As I look at the study in detail, I see that most of the combination group indeed did not maintain the bupropion beyond the 3 month mark. I suspect that if people were strongly encouraged to continue the treatments longer, on a preventative basis, then the abstinence rates could have been much higher than 35%.

Here is a 2005 meta-analysis showing that the tricyclic antidepressant nortriptyline can be effective. Once again, the effects were significant but modest. Most of the studies used only a standard 3-month course of treatment, followed by a taper and discontinuing the nortriptyline. In the one study allowing a full year of nortriptyline treatment, the abstinence rate was much higher (40%):
http://www.ncbi.nlm.nih.gov/pubmed/15733245?ordinalpos=1&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DiscoveryPanel.Pubmed_Discovery_RA&linkpos=4&log$=relatedreviews&logdbfrom=pubmed

Similarly, in a study maintaining varenicline for 52 weeks, the abstinence rate was 36.7%, compared to 7.9% with placebo. However, while the existing evidence about the safety of using varenicline on a long-term basis is generally reassuring, more long-term experience is necessary with this drug to know for sure. I think the potential risks would have to be weighed against the risks of continuing to smoke. Here is a link to the study:
http://www.ncbi.nlm.nih.gov/pubmed/17407636?ordinalpos=1&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DiscoveryPanel.Pubmed_Discovery_RA&linkpos=1&log$=relatedarticles&logdbfrom=pubmed

Similarly again, in the one long-term study of bupropion (a full year of medication), there were considerably higher abstinence rates:
http://www.ncbi.nlm.nih.gov/pubmed/11560455?ordinalpos=18&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DefaultReportPanel.Pubmed_RVDocSum

The following small study showed that varenicline helped reduce smoking in patients with schizophrenia, and appeared to have some beneficial cognitive effects in this group.
http://www.ncbi.nlm.nih.gov/pubmed/19251401?ordinalpos=6&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DefaultReportPanel.Pubmed_RVDocSum

The following small study showed possible increased abstinence rates when varenicline and bupropion therapy was combined:
http://www.ncbi.nlm.nih.gov/pubmed/19246427?ordinalpos=8&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DefaultReportPanel.Pubmed_RVDocSum

Another study supporting the idea that combination therapy (e.g. varenicline + nicotine replacement) is more effective than one treatment alone, for helping smokers quit:
http://www.ncbi.nlm.nih.gov/pubmed/18826906?ordinalpos=56&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DefaultReportPanel.Pubmed_RVDocSum

The following study shows that varenicline is similarly tolerated and effective in patients with depression, compared with patients without a history of depression. Stress and mood scores improved slightly with time:
http://www.ncbi.nlm.nih.gov/pubmed/19238488?ordinalpos=12&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DefaultReportPanel.Pubmed_RVDocSum

This 2009 study from Biological Psychiatry suggests that varenicline could also reduce alcohol consumption in heavy-drinking smokers:
http://www.ncbi.nlm.nih.gov/pubmed/19249750?ordinalpos=2&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DefaultReportPanel.Pubmed_RVDocSum

This is a 2009 Cochrane review of smoking relapse prevention interventions; it supports extended treatment with varenicline to prevent relapse, and concludes that there is insufficient evidence at this point to comment one way or another on specific behaviour therapies:
http://www.ncbi.nlm.nih.gov/pubmed/19160228?ordinalpos=1&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DiscoveryPanel.Pubmed_Discovery_RA&linkpos=1&log$=relatedreviews&logdbfrom=pubmed

This study looked at 20 weeks of adjunctive CBT, and found no significant difference in abstinence rates after a year. But it did find an advantage in the CBT group in the shorter term, during the course of CBT (45% abstinence in the CBT group vs. 29% in the control group, at the 20 week mark). This suggests that long-term, ongoing, continuous CBT may be helpful to boost abstinence rates, but the therapy loses its effectiveness if it is not maintained:
http://www.ncbi.nlm.nih.gov/pubmed/18855829?ordinalpos=3&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DefaultReportPanel.Pubmed_RVDocSum

All of these studies support the idea that smoking addiction is a long-term problem. Short-term strategies (typically over a few months) definitely help, but long-term, continuing effort or treatment is needed to maintain abstinence for most people. These strategies could include medications such as varenicline, bupropion, or nortriptyline; and they could include psychotherapeutic approaches such as CBT.

Individuals with psychiatric illnesses such as depression, bipolar disorder, ADHD, and especially schizophrenia, have much higher rates of smoking. Here is a reference:
http://www.ncbi.nlm.nih.gov/pubmed/15949648?ordinalpos=1&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DiscoveryPanel.Pubmed_Discovery_RA&linkpos=2&log$=relatedarticles&logdbfrom=pubmed

There is evidence that nicotine can acutely improve elements of cognitive performance and to reduce impulsivity, particularly in those with illnesses such as schizophrenia and ADHD. This may be one of the reasons why individuals with these problems are more drawn to cigarette smoking. Here is some evidence:
http://www.ncbi.nlm.nih.gov/pubmed/17443126?ordinalpos=6&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DefaultReportPanel.Pubmed_RVDocSum

http://www.ncbi.nlm.nih.gov/pubmed/18022679

Also there is evidence that nicotine can improve performance in attention tests in elderly people with dementia:
http://www.ncbi.nlm.nih.gov/pubmed/10326778?ordinalpos=1&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DiscoveryPanel.Pubmed_Discovery_RA&linkpos=3&log$=relatedarticles&logdbfrom=pubmed

Yet, of course, nicotine has numerous harmful effects. And it is likely that nicotine could cause long-term harm to cognitive function, through several mechanisms, even if it causes short-term enhancement. A medication such as varenicline, due to its agonist effect on nicotine receptors, may be especially helpful to address some of the cognitive or attentional problems in persons with mental illnesses.

In terms of health care policy, I am puzzled about why effective therapies to improve smoking cessation are not publicly funded. Smoking is one of the largest public health problems in the world, and causes an enormous burden of premature disease and death, as well as an enormous financial drain on the health care system. I believe that all proven therapies for smoking cessation should be freely available.

Unfortunately, varenicline -- and other anti-smoking therapies -- are expensive, and they are often not covered by health plans.