A 51-Year-Old Woman With Bipolar Disorder Who Wants to Quit Smoking

DR TESS: Ms G is a 51-year-old woman with a history of bipolar disorder and tobacco use for more than 35 years. She has tried to quit smoking for 8 years and is seeking specialist input as to how to proceed.

Ms G began smoking as a teenager. At age 43 years she first attempted to quit using the nicotine patch and hypnosis therapy, but was only able to cut down to 4 cigarettes a day. Since then she has tried other types of nicotine replacement therapy with limited success. She enrolled in group therapy, during which she stopped smoking transiently, but returned to smoking after the sessions ended. Most recently she was prescribed varenicline but was very reluctant to start because of concerns about adverse effects. After discussing the medication with her primary care physician, she began taking it 6 weeks ago. Since then she has been able to again reduce her use to 4 cigarettes a day; however, she believes she is not ready to set a quit date.

Ms G was diagnosed as having bipolar disorder at age 36 years. Since then she has had several periods of debilitating depression requiring 4 hospitalizations. Her mood disorder has been treated with lithium and carbamazepine in the past. She is currently maintained on citalopram and risperidone. She finds that periods of unstable mood are the most difficult times to stop smoking. She is currently doing well and reports that her psychiatric symptoms are well controlled.

Her medical history is otherwise notable for probable chronic obstructive pulmonary disease, diagnosed on the basis of findings of emphysema on a chest radiograph and recurrent bronchitis. She has had atypical chest pain in the past, with a negative exercise stress test result in 2003. She is postmenopausal and received hormone therapy in the past. Although her bone density scan result was normal 5 years ago, her most recent scan showed evidence of osteopenia. She also has anemia of chronic disease. Her Papanicolaou test and mammogram results have been normal.

Her medications include varenicline, 1 mg twice daily; bupropion, 150 mg sustained-release twice daily; gabapentin prescribed for mood stabilization, 400 mg in the morning and 800 mg at night; risperidone, 1 mg at night; and citalopram, 20 mg daily. She has no known drug allergies.

Ms G has 3 siblings and graduated from college. She lives alone in elder and disabled housing and has limited social support. She has a remote history of drug use but not intravenous drug use. She has a history of heavy alcohol use but none in more than 20 years. She currently volunteers in the community and is looking for permanent employment.

Her father died 1 year ago; her mother is in her 80s and is healthy. Her maternal grandmother and aunt have been diagnosed as having diabetes. There is a history of bipolar disorder in a paternal aunt.

I can't really remember how I started smoking. Somebody offered me a cigarette and I tried it. I really did like it when I started smoking. It's become a way to relax.

I first tried to quit in 1999. I was on the patch, and I went to a group for self-hypnosis. It was hard because the woman running the group wanted us to cut down instead of stopping immediately. I knew I could have a cigarette only at certain hours. So in between, I was constantly thinking about having a cigarette. I was obsessed with it. Quitting is anxiety-provoking. You feel like you're getting your arm cut off, like you're going to miss that whole part. That's another thing I don't like about cigarettes. They have that control over you.

The second time I tried to quit, I went through an educational program at a hospital that was really helpful. I had a one-on-one person I saw once a week. I was also doing it with other people. When the program ended, I thought, “Oh, I’ll just have 1.” I was having 3 cigarettes a day for a while and then, gradually, more and more. And it was a pack a day again.

In 2000, I was doing the patch but then I was smoking while on the patch. And I thought, “I’m going to have a heart attack, so I’d better not do that.” Then I started chewing the gum, but my teeth are not very good, so it was hard to chew. In 2006, I was using the lozenges and feeling pretty good. But I was having fewer cigarettes and more lozenges, so I wasn't really cutting down on the nicotine.

I started medication [varenicline] about 2 months ago, and I’ve gradually been cutting down. I don't feel like I want to smoke as much. After the ninth day, you're supposed to quit totally, but I haven't done that. I’ve just been smoking 4 cigarettes a day. I am committed to stopping smoking, but I just keep putting off the day. I’m having trouble setting the date because I guess I’m not full-hearted into quitting.

I am wondering if Dr Schroeder thinks I’m going to have to keep taking [varenicline]. I know it's connecting certain things in my brain, but are the connections going to stay that way once I stop taking the medications?

What is the epidemiology of smoking in patients with mental illness in general, in the United States, and elsewhere? What is known about the pathophysiology of smoking and its impact on general health and quality of life in this population? Does smoking cessation exacerbate chronic mental illness? What treatment strategies are available, when are they indicated, and what is the evidence that they are effective in this population? What do you recommend for Ms G?

DR SCHROEDER: Ms G, a woman with bipolar disorder who wishes to quit smoking, also represents the larger public health question of smoking cessation in patients with mental illness. Although few trials of smoking cessation in patients with mental illness have been conducted, the serious cardiovascular consequences and increased mortality associated with smoking make smoking cessation essential to incorporate in the treatment of patients with mental illness. Nevertheless, a number of issues specific to treating patients with mental illness have raised clinicians' concerns about smoking cessation. Herein, I examine the evidence that addresses these concerns.

Ms G brings many assets to her desire to quit smoking: interest in her health as manifested by attendance at a gym, awareness of the respiratory complications of smoking for her and her family, previous quit attempts, her current low smoking level of only 4 cigarettes per day, the stability of her bipolar disorder, her tolerance of varenicline, which has diminished her cravings for nicotine, and her desire to have smoking cessation counseling. She also faces challenges: a 35-year history of smoking, the complications of her mental illness and its attendant medications, her ambivalence about stopping, and the stigma associated with both smoking and mental illness.

Smoking is probably the most common factor contributing to the deaths of persons with mental illness and/or substance abuse disorders, yet this has been a “silent” epidemic.^{1 - 3} Rates of smoking are much higher in all categories of mental illness and substance abuse than in the general US population (Table 1).^{4 - 11} The US Centers for Disease Control and Prevention currently report a record low smoking prevalence rate of 19.8%¹¹ in the general population, which would be even lower if it excluded persons with mental illness and substance abuse.

Despite constituting only 22% of the population, persons with mental illness including substance abuse disorders consume an estimated 44%⁵ of all cigarettes sold in the United States.^{12 - 13} This heavy consumption reflects higher smoking prevalences and the fact that these patients tend to smoke more cigarettes.^{5 ,14} Forty-one percent of current smokers report having a mental health diagnosis in the past month and 60% report a mental health diagnosis at some point in their lifetime.¹² Among current smokers, the most common mental health diagnoses (reported in the past 30 days) are alcohol abuse, major depressive disorder, anxiety disorders, and illicit drug use or dependence.¹² The more severe the psychiatric symptoms, the more likely the patient is to be a smoker.¹⁵

These smoking rates take a substantial toll on the population of individuals with mental illness. Although exact numbers are unknown, perhaps as many as 200 000 of the 435 000 annual deaths from smoking in the United States occur among persons with mental illness and/or substance abuse.^{1 ,16 - 17} People with chronic mental illness die, on average, 25 years earlier than the general population,^{3 ,17} largely from cardiovascular disease and diabetes mellitus.³

For people with a history of alcohol abuse, such as Ms G, smoking also takes a huge toll. Quiz Ref IDIn a 20-year follow-up study of 845 patients hospitalized for addiction to alcohol or illicit drugs, the observed mortality was 48% vs an expected rate of 18%; half the deaths were attributable to smoking.¹⁸ In addition to morbidity and mortality, smoking exerts other tolls: persons with mental illness are more likely to have fixed budgets, and cigarettes may consume more than 25% of the incomes of those receiving public assistance.^{19 - 20} Finally, the odor and visual signs of smoking add to the stigma of mental illness, thereby affecting social interaction and job seeking—an activity currently consuming much of Ms G's energy.^{21 - 23}

In psychiatric hospitals and clinics, cigarettes have historically been used as rewards or punishment, with smoking breaks permitted as a privilege.^{24 - 26} Though the practice is waning, it is still in place in some institutions.²⁷ The National Association of State Mental Health Program Directors Research Institute Inc surveyed all state psychiatric hospitals in 2006, with an 82% response rate. Among responding hospitals, 41% reported being nonsmoking and 59% permitted smoking by patients on hospital premises.²⁷ Patient permission to smoke was based on privilege status in the 34% of respondents that permit smoking. Quiz Ref IDThe report further stated that many hospitals responded that gaining access or permission to smoke was a motivator for patients to comply with staff. Family members and staff often viewed smoking as one of the patients' few pleasures and were reluctant to eliminate it. The tobacco industry has been implicated in sponsoring studies that claim that nicotine helps psychiatric symptoms and in opposing smoking bans in institutions.²⁸

The enormity of this disease burden, coupled with the emerging data on the dangers of secondhand smoke, finally galvanized attention about smoking in this population and emphasized the importance of integrating and co-locating psychiatric and medical services, especially smoking cessation.²⁴ The National Association of State Mental Health Program Directors Council has issued a call to action²⁶ and the National Mental Health Partnership for Wellness and Smoking Cessation, a coalition of more than 30 organizations representing mental health clinicians, advocacy, and governmental groups, pledged to promote smoking cessation strategies in settings where mental illness is treated and to change the culture surrounding smoking and mental illness.²⁹

Among the more than 4800 active compounds in tobacco smoke, the addictive component is nicotine. Nicotine stimulates the release of multiple neurotransmitters, including dopamine, norepinephrine, acetylcholine, glutamate, serotonin, β-endorphin, and γ-aminobutyric acid.^{30 - 31} Of particular importance for smokers is nicotine's activation of the dopamine reward pathway.³⁰ This pathway—located in the midbrain—is stimulated by crucial biological behaviors that confer species survival value, such as eating when hungry, drinking when thirsty, and engaging in sex.³⁰ Nicotine also helps smokers to concentrate and may help to regulate mood.³⁰ Within seconds after inhaling a cigarette, nicotine enters the brain and stimulates dopamine release and the resultant feeling of pleasure.³¹ Some authors have considered nicotine addiction a chronic brain disease, and neuroimaging studies of long-term smokers show up-regulation of nicotine receptors in specific brain regions.^{32 - 33} In addition, smoking may result in neuroplastic changes that persist for months to years.³⁰

Tobacco use and dependence is highly heritable. Genetically slow metabolism of nicotine may be associated with lower levels of dependence.³⁰ Specific genetic patterns predispose to both depression and nicotine dependence.^{34 - 36} Genes that alter dopamine function and transmission may influence the reinforcing effects of smoking, as well as tobacco dependence and the risk of relapse.³⁷ In adolescents, the presence of the DRD2 A1 dopamine receptor allele influences the probability of smoking progression.³⁸

Biological factors also contribute to the high rates of smoking in mental illness, but the “self-medication” hypothesis—that persons with depression smoke to reduce depressive symptoms—has little empirical evidence to support it.^{39 - 40} For example, the dopamine release mediated by nicotine may reduce some symptoms in schizophrenia, thereby creating an incentive to smoke.⁴¹ Other chemicals in smoke may reduce the symptoms of depression by reducing the monoamine oxidase (MAO) enzymes MAOA and MAOB, similar to the effects of MAO inhibitor antidepressants.^{30 ,42} Nicotine also creates a general arousal reaction, which might counter the sedating effects of some psychotropic medications.³⁰ Quiz Ref IDHowever, when dependent smokers are deprived of nicotine, they develop withdrawal symptoms, including depression, insomnia, anxiety, irritability, frustration, anger, difficulty concentrating, restlessness, increased appetite, and weight gain.⁴³ Ms G describes these withdrawal symptoms with her prior quit attempts. Because these symptoms mirror those of psychiatric illnesses, especially depression, it is challenging to treat a depressed patient who is trying to stop smoking. Finally, tobacco users on stable regimens of antipsychotic medications and some antidepressants may experience adverse drug events when withdrawing from smoking because of changes in the metabolism of their medication, related to the effects of nicotine on the cytochrome P450 enzyme system.⁴⁴ For example, smokers have 24% lower levels of phenothiazines, 70% lower levels of haloperidol, and 50% lower levels of caffeine than when they are not smoking.⁴⁴ When the smoking level is constant, drug metabolism is also constant, but when smoking is either decreasing or increasing, it can result in either unduly high or low medication levels.

Some mental health clinicians have traditionally been reluctant to encourage their patients to stop smoking for fear that it will worsen the underlying psychiatric condition. While there is little evidence for this assumption, the association is difficult to disprove.²⁴ The constellation of symptoms that occur during nicotine withdrawal closely parallels those of depression^{45 - 46} and the waxing and waning course of mental illness may be difficult to distinguish from the symptoms of withdrawal. Published studies have offered conflicting evidence, with some finding that relapse of depression is related to smoking cessation^{46 - 48} and others finding no association.^{49 - 50} The tobacco industry has supported some of the studies that concluded that cessation worsens mental health, raising questions about the impartiality of the research.²⁸ One prospective study finding an association between cessation and mental illness controlled for a history of depression and reported a 7-fold increased risk of relapse of depression following smoking cessation, but 39% of the relapsed smokers in the study were lost to follow-up; thus, their depression status could not be confirmed.⁴⁷ In contrast, a prospective study of 304 smokers in cessation treatment, 32% of whom had a history of depression, found no relation between cigarette abstinence or reduction and depressive episodes after treatment for smoking cessation and concluded that depression recurrence was most strongly predicted by depression history, not by smoking status.⁴⁹ A randomized trial of 322 actively depressed smokers found no differences in improvement in depression between those who successfully stopped smoking and those who did not and concluded that individuals in treatment for clinical depression can successfully undertake cessation treatment without adverse mental health effects.⁵⁰

One natural experiment indicates that stopping smoking may not destabilize psychiatric conditions. Quiz Ref IDPsychiatric hospitals^{17 ,24 ,26} are now becoming smoke free, despite initial resistance from staff⁵¹ —many of whom smoke⁵² —as well as from patients and families²⁴ and from the tobacco industry.^{28 ,53} Some staff had warned that removing tobacco privileges would lead to increased violence, disciplinary actions, and exacerbations of psychiatric symptoms. In fact, the opposite occurred, with less violence,²⁶ fewer disciplinary actions,^{24 ,52} and more staff contact time with patients in therapeutic settings being reported.^{24 ,26}

Smoking cessation interventions comprise counseling and pharmacological agents. Tobacco use counseling treatment may be delivered in a variety of formats, including proactive telephone counseling or contact (quit lines and call-back counseling), individual counseling, and group counseling/contact. All forms of intervention increase abstinence rates compared with no intervention.⁵⁴ A Cochrane review analyzed the effects of smoking cessation programs delivered in a group format compared with self-help materials or with no intervention, compared the effectiveness of group therapy and individual counseling, and determined the effect of adding group therapy to advice from a health care professional or to nicotine replacement. The study also determined whether specific components increased the effectiveness of group therapy. Sixteen studies compared a group program with a self-help program, and cessation increased with the use of a group program (n = 4395; odds ratio [OR], 2.04; 95% confidence interval [CI], 1.60-2.60). Group programs were more effective than no intervention (7 trials; n = 815; OR, 2.17; 95% CI, 1.37-3.45). Group therapy was not more effective than a similar intensity of individual counseling, but the addition of group therapy to other forms of treatment, such as advice from a health care professional or nicotine replacement, appeared to produce extra benefit.^{55 - 56}

Pharmaceutical interventions include 5 versions of nicotine replacement therapy—a long-acting transdermal patch and 4 shorter-acting formulations (gum, lozenge, nasal spray, and inhaler)—plus 2 drugs that act on the central nervous system, bupropion and varenicline (Table 2).^{54 ,57 - 59} Bupropion is an antidepressant that also reduces cravings for nicotine.⁶⁰ Varenicline, which was approved by the US Food and Drug Administration (FDA) in 2006 for smoking cessation, is a partial nicotinic cholinergic receptor agonist.^{58 - 61} The efficacy of these medications has been extensively reviewed in the Treating Tobacco Use and Dependence: 2008 Update Practice Guideline.⁵⁴ The update rates the strength of evidence for its recommendations to be level A (“multiple, well-designed randomized clinical trials, directly relevant to the recommendation, yielded a consistent pattern of findings”) for all 7 of these FDA-approved first-line smoking cessation medications. The antidepressant nortriptyline has efficacy similar to bupropion for smoking cessation⁶¹ but is seldom used for smoking cessation.⁵⁴ The efficacy of bupropion and nortriptyline is unrelated to their antidepressant effects. Clonidine is categorized as a second-tier drug but also is seldom used for smoking cessation.⁵⁴ Selective serotonin reuptake inhibitors, which are commonly prescribed in mental illness, have not been shown to reduce smoking rates.^{61 - 62} Opiate antagonists have been tried for smoking cessation, but according to a Cochrane review, there is insufficient evidence to determine whether naltrexone is effective for this indication.⁶³

Multiple trials have documented that counseling alone, medication alone, and the combination of counseling and medication all improve cessation rates compared with placebo.^{43 ,54 ,56} Smoking cessation efficacy rates at 5 months or more after study initiation are as follows. In terms of behavioral interventions, physician advice to quit has an estimated OR of 1.3 (95% CI, 1.1-1.6) compared with no advice to quit, with a percentage abstinence of 10.2% (95% CI, 8.5%-12.0%) vs 7.9%, respectively. Physician counseling provides an OR for abstinence of 2.2 (95% CI, 1.5-3.2), vs 1.7 (95% CI, 1.3-2.1) for counseling by a nonphysician compared with no counseling. Abstinence rates for counseling are 19.9% (95% CI, 13.7%-26.2%), 15.8% (95% CI, 12.8%-18.8%), and 10.2%, respectively. Among types of smoking cessation counseling, individual is most effective (OR, 1.7 [95% CI, 1.4-2.0]; abstinence, 16.8% [95% CI, 14.7%-19.1%]), followed by group (OR, 1.3 [95% CI, 1.1-1.6]; abstinence, 13.9% [95% CI, 11.6%-16.1%]), proactive telephone counseling (OR, 1.2 [95% CI, 1.1-1.4]; abstinence, 13.1% [95% CI, 11.4%-14.8%]), and self-help (OR, 1.2 [95% CI, 1.0-1.3]; abstinence, 12.3% [95% CI, 10.9%-13.6%]), all of which are comparable but all more effective than no intervention (abstinence rate, 10.8%).

In terms of pharmacotherapy, the first-line agents (compared with a placebo plus counseling control group, which had an abstinence rate of 13.8%) include varenicline (2 mg/d) (OR of quitting of 3.1 [95% CI, 2.5-3.8]; abstinence, 33.2% [95% CI, 28.9%-37.8%]); bupropion sustained-release (OR, 2.0 [95% CI, 1.8-2.2]; abstinence, 24.2% [95% CI, 22.2%-26.4%]), nicotine nasal spray (OR, 2.3 [95% CI, 1.7-3.0]; abstinence, 26.7% [95% CI, 21.5%-32.7%]); nicotine inhaler (OR, 2.1 [95% CI, 1.5-2.9]; abstinence, 24.8% [95% CI, 19.1%-31.6%]); nicotine lozenge (2 mg) (OR, 2.0 [95% CI, 1.4-2.8]; abstinence, 24.2% [95% CI not reported]); nicotine patch (for 6-14 weeks) (OR, 1.9 [95% CI, 1.7-2.2]; abstinence, 23.4% [95% CI, 21.3%-25.8%]); and nicotine gum (for 6-14 weeks) (OR, 1.5 [95% CI, 1.2-1.7]; abstinence rate, 19.0% [95% CI, 16.5%-21.9%]). Second-line agents (not approved for labeling for this indication by the FDA) include clonidine (OR, 2.1 [95% CI, 1.2-3.7]; abstinence, 25.0% [95% CI, 15.7%-37.3%]) and nortriptyline (OR, 1.8 [95% CI, 1.3-2.6]; abstinence, 22.5% [95% CI, 16.8%-29.4%]). The relative ORs derive from individual studies and do not allow for cross-comparisons of the different pharmacologic agents.

Combination therapy includes the nicotine patch (>14 weeks) plus ad lib nicotine (gum or nasal spray), with an OR of 3.6 (95% CI, 2.5-5.2) and an abstinence rate of 36.5% (95% CI, 28.6%-45.3%); nicotine patch plus bupropion sustained-release (OR, 2.5 [95% CI, 1.9-3.4]; abstinence rate, 28.9% [95% CI, 23.5%-35.1%]); nicotine patch plus nortriptyline (OR, 2.3 [95% CI, 1.3-4.2]; abstinence, 27.3% [95% CI, 17.2%-40.4%]); and nicotine patch plus nicotine inhaler (OR, 2.2 [95% CI, 1.2-3.6]; abstinence, 25.8% [95% CI, 17.4%-36.5%]).^{54 ,57}

Toll-free telephone quit lines—now available for smokers in every state—increase quit rates in a range comparable with drug trials.^{64 - 66} Evidence for the efficacy of online quit programs, some but not all of which charge for their services, is also favorable, although less robust than for quit lines.⁶⁷ Multi-impact systems such as those used by Kaiser Permanente of Northern California have successfully reduced smoking rates in their populations. Using a variety of interventions, including aggressive case-finding plus referral to cessation experts and quit lines, Kaiser was able to reduce its adult smoking prevalence from 12% to 9% in just 3 years.²⁹

Several caveats apply to studies of smoking cessation. First, although smoking changes brain neurotransmitters and neuroplasticity—which should support categorizing it as a chronic disease—the therapies tested in most drug trials stop after 12 weeks.⁵⁴ In contrast, treatment for other addictions, such as methadone maintenance for heroin addiction, is provided over a much longer duration. It is possible that longer treatment could yield higher cessation rates. For example, 52 weeks of nortriptyline plus extended counseling resulted in 50% cessation rates at the end of treatment.⁶⁸ Second, smoking is a heterogeneous disease with differing degrees of use and addiction. It may be difficult to generalize results from 1 study to a given patient. Third, the published smoking cessation trials may be unduly optimistic for the average clinic population; quit rates in both the treatment and placebo groups of drug trials are higher than in the general population because of the high motivation of study participants and the higher quality, frequency, and duration of the counseling sessions. Spontaneous unaided quit rates are probably about 4%.⁵⁴ Also, since the circumstances of each trial are different, one should be wary of making generalizations as to the relative efficacies of the different smoking cessation agents. Finally, drug trials typically exclude patients with mental illness, thereby eliminating a large number of smokers who may have lower cessation rates.

Although smoking cessation rates are generally considered to be lower in persons with mental illness and/or substance abuse,⁴⁸ there is abundant evidence that smokers with mental illness are able to quit. About 70% of smokers with mental illness want to quit,^{54 ,69 - 71} similar to the general population.⁷² Quiz Ref IDFurthermore, a review of 24 published studies for all mental illnesses⁷³ and a meta-analysis of 15 reports on patients with a history of major depression⁷⁴ both concluded that rates of cessation among these patients were comparable with those of the general population. Specifically for major depression, studies have documented 12-month cessation rates of 22% using the combination of bupropion and counseling⁷⁵ and as high as 32% for intensive counseling without pharmacotherapy.⁷⁶ Therefore, clinicians should be encouraged to treat patients with depression and other forms of mental illness, such as Ms G, who want and are able to quit smoking. Unfortunately, clinicians, and physicians in particular, frequently do not incorporate smoking cessation into treatment for their patients with mental health diagnoses.^{77 - 78} A 2007 study conducted by the Association of American Medical Colleges/American Legacy Foundation compared physician practice patterns with respect to smoking cessation and found that while 89% of internists and 62% of psychiatrists ask their patients about smoking, only 19% and 14% arrange follow-up and 7% and 2%, respectively, refer patients to quit lines.⁷⁹ Another study found that 60% of nurses surveyed believed that nicotine was the component in cigarettes that causes cancer and, therefore, were reluctant to recommend nicotine replacement therapy.⁸⁰ In my experience, other clinicians have similar misunderstandings.

Standard cessation treatments of counseling and pharmacotherapy are appropriate in the mentally ill population.⁵⁴ Bupropion should be considered for smokers with chronic depression, as it could have a double effect.

Studies to date have shown higher cessation rates for varenicline plus counseling than for bupropion plus counseling,^{58 - 59 ,81 - 82} all but 1 of which⁸² have excluded patients with mental illness. There have been reports of suicidal ideation and worsening psychiatric symptoms in patients taking varenicline, including some who have continued to smoke.^{83 - 85} The FDA issued an alert and required Pfizer to provide a warning for varenicline (Chantix) and a change to its label and medication guide to state that “[s]ome patients have reported changes in behavior, agitation, depressed mood, suicidal thoughts or actions when attempting to quit smoking while taking Chantix or after stopping Chantix.” It states that patients experiencing such changes should stop taking Chantix and contact their physician.

Of note, bupropion also carries such a warning (http://us.gsk.com/products/assets/us_zyban.pdf). The association between varenicline use and exacerbation of mental illness, the frequency of which has yet to be ascertained, must be balanced against the very high risk of continued smoking. Drug safety trials of varenicline among patients with mental illness are currently recruiting study participants (available at http://www.clinicaltrials.gov by using a search string for varenicline and schizophrenia or mental illness).

Arguably, for a patient with acute psychiatric symptoms, cessation attempts should be deferred until they are stabilized,² although much more research is needed on appropriate timing.

There are many unanswered questions about timing of cessation attempts for those with mental illness, long-term efficacy of treatments, and harm-reduction strategies that focus on decreasing the number of cigarettes smoked, rather than cessation. Perhaps most importantly, what will it take to command clinicians' attention to the devastating hidden epidemic in this population and undertake cessation efforts to stem it?

Answers to these questions will come sooner if the mental health treatment community and the clients and families they represent advocate more strongly in the future than they have in the past. No other issue is as important for the long-term health and well-being of mentally ill patients.

For individual patients, the question becomes how can we better motivate the mental health treatment community to give smoking and mental illness a higher priority?⁸⁶ What is the role of non–mental health clinicians, such as internists and family physicians, in trying to help patients with mental illness stop smoking?

As desirable as it might be to transform primary care and all other relevant clinicians—pulmonary specialists, diabetologists, cardiologists, obstetricians—into smoking cessation experts, it is an unrealistic goal. The achievable alternative, and one that I challenge and charge all clinicians with attaining, is to develop systems and resources to marshal essential nicotine cessation services. At a minimum, this should include referring the smoker to a quit line.⁷⁷ No smoker should leave a physician's office without an offer of help to stop smoking.

Ms G should be congratulated for reducing her smoking level to 4 daily cigarettes. Now she should set a date to quit entirely. She is already taking 2 recommended smoking cessation drugs, bupropion and varenicline. There are as yet no studies assessing the efficacy of varenicline in combination with other cessation drugs, although anecdotally, such off-label use is not uncommon. The best cessation results come from a combination of counseling and drugs,⁵⁴ and Ms G has indicated that she would welcome such counseling. If no counseling services are available at her institution, she should be directed to the free telephone quit line services (1-800-QUIT NOW). This system refers callers to their own state's quit line via this national routing number. Counseling should emphasize the continuing benefits of quitting as well as devise behavioral strategies to combat the urge to smoke. As to her concern about the duration of varenicline treatment, 1 published report has shown better cessation rates from 24-week therapy as opposed to the standard 12 weeks.⁸¹ The FDA has approved use of the drug for 6 months for those who initially responded to the drug. Another study of 251 patients randomized to varenicline and 126 to placebo demonstrated that patients can safely use the drug for up to 52 weeks.⁸⁷ Approximately half of the participants (53.8% in the varenicline group and 46.8% in the placebo group) completed the study. There are no population-based data yet regarding safety or possible adverse effects from longer-term use. Of paramount importance to Ms G and other patients with mental illness is optimal treatment and careful monitoring of her psychotropic medications, as well as caution about symptoms of caffeine excess—both conditions of concern in the setting of decreased tobacco use. Her psychiatrist and internist should also be alert to the serious but rare complications that may be associated with varenicline use: suicidal ideation and mania. Should her current cessation program fail, Ms G should be offered another effort using both short- and longer-acting nicotine replacement therapy—with or without bupropion.

QUESTION: You spoke only briefly about weight gain, but in my practice, that is a tremendous issue for patients who contemplate quitting smoking. Are there any particular strategies for preventing weight gain in this scenario?

DR SCHROEDER: Weight gain is a usual consequence of quitting smoking, whether one has mental illness or not, and whether or not cessation therapies are used in the quit attempt.⁵⁴ The weight gain is said to average about 8 lb but it can be more. Many of the standard psychotropic medications are also associated with weight gain, so this can be an important compounding issue for many people with mental illness. Using bupropion in concert with nicotine replacement therapy, particularly the 4-mg nicotine gum and the lozenges, may delay weight gain, but it still occurs.^{54 ,57} It is probably best not to try to get people to change their diet at the same time they are stopping smoking, but once smoking is conquered, weight reduction strategies might be appropriate. It is also worth trying to couple cessation with increased exercise, instituting this in simple steps, such as advising patients to park at the far end of a parking lot instead of the closest spot.

QUESTION: The Department of Veterans Affairs (VA) has been very aggressive in trying to reduce the rampant smoking throughout its facilities. Are there data on outcomes in terms of success or failure rates, changes in the costs of caring for the population, and rates of tobacco-associated diseases?

DR SCHROEDER: The tobacco industry used to send cigarettes to soldiers in combat. During World War I, General Pershing said, “Send us two things to beat the Germans, bullets and cigarettes.” So we induced young men to smoke. The VA has recently taken on tobacco use as a major issue.⁸⁸ The VA's facilities actually track cessation efforts as a medical quality indicator and rate their physicians accordingly.⁸⁹ As a result, the VA does better than most systems in asking about smoking and urging cessation. Smoking cessation rates have not gone down as much as the VA would like, but they are declining.⁹⁰ And VA facilities are now using telephone quit lines as a successful cessation option.⁹¹ Many veterans live at some distance from the local VA, so for them quit lines are a preferred option. I am not aware of whether there have been any documented declines in smoking-related illnesses among veterans as a result of these efforts.

MS G’S PRIMARY CARE PHYSICIAN: Ms G is receiving ongoing mental health care. She is seeing a psychiatrist and a social worker regularly. When you refer to her needing “counseling,” do you mean, specifically, counseling directed at smoking cessation?

DR SCHROEDER: Psychiatrists and psychiatric social workers are not limited to the typical brief encounter with patients that primary care physicians are. So they have more time to splice in smoking cessation counseling (cognitive and behavioral therapy) along with standard psychotherapy. But they typically don't do it. Why? First, they don't think of it because it is not yet part of the mental health treatment culture. Second, they don't feel comfortable doing it because they lack the training and expertise or, perhaps, because they are smokers themselves. Our center is attempting to make mental health clinicians more knowledgeable about smoking cessation strategies by providing them with a tailored version of the cessation curriculum Rx for Change, prepared by pharmacy faculty at the University of California, San Francisco (UCSF).⁵⁷ This curriculum demonstrated significant changes in practice behavior of psychiatric residents with respect to undertaking cessation efforts, validated through chart review.⁹² We will help disseminate that curriculum so counselors will feel more comfortable with cessation efforts.

QUESTION: Can you comment on the role of restrictions in public smoking in the decreasing prevalence of tobacco use?

DR SCHROEDER: Credit for the decline in smoking prevalence over the past 4 decades generally focuses on a set of policy measures, especially tobacco taxes, clean indoor air laws, countermarketing, and smoking cessation efforts.⁹³ Smoking is becoming denormalized and increasingly stigmatized.^{22 - 23} It is disappointing that the prevalence of smoking has plateaued recently, after years of progressive steady decline. But the Centers for Disease Control and Prevention recently reported that smoking prevalence for the first 6 months of 2007 was at a modern age-adjusted low of 19.8%.¹¹ There are tremendous variations in population smoking prevalence. Smoking is much more prevalent in southern states, among the poor, and in people with low educational attainment.⁷⁷ A recent study showed that patients resume smoking after discharge from mental hospitals but also commented that they were not given adequate cessation assistance.⁹⁴

QUESTION: If the National Institutes of Health (NIH) is designed to improve the health of Americans and if the biggest preventable problem is tobacco, do you think the NIH ought to change the way it is organized?

DR SCHROEDER: The reality today is that the NIH is more concerned with basic mechanisms of disease and molecular biology than in researching the efficacy of various clinical strategies. The latter role tends to be financed by the pharmaceutical industry, but since so many of the evidence-based cessation techniques are either off-patent drugs or counseling, there is little incentive for industry to study these or to analyze the efficacy of various combinations. So, there needs to be pressure on the NIH to expand its research agenda on smoking cessation. But pressures of that sort require well-organized and passionate advocacy groups of patients, families, and clinicians. Because of the stigma surrounding smoking, there are no such advocacy groups. It may be up to the medical and public health community to exert that pressure.¹