|Home | About | Journals | Submit | Contact Us | Français|
Despite recent advances in understanding the causes and treatment of nicotine dependence among individuals with psychiatric disorders, smoking among individuals with bipolar disorder (BD) has received little attention. The goal of this review is to synthesize the literature on the epidemiology, consequences, and treatment of smoking and nicotine dependence among individuals with BD and to delineate a future research agenda.
We conducted a PubMed search of English-language articles using the search terms “bipolar disorder,” “mania,” “tobacco,” “nicotine,” “and “smoking,” followed by a manual search of the literature cited in the identified articles. Articles were chosen by the authors on the basis of their relevance to the topic areas covered in this selective review.
Adults with BD are 2 to 3 times more likely to have started smoking and, on the basis of epidemiological data, may be less likely to initiate and/or maintain smoking abstinence than individuals without psychiatric disorders. Smoking cessation is achievable for individuals with BD, but challenges such as chronic mood dysregulation, high prevalence of alcohol and drug use, more severe nicotine dependence, and limited social support can make quitting more difficult. Effective treatments for tobacco cessation are available, but no controlled trials in smokers with BD have been conducted.
Cigarette smoking is a prevalent and devastating addiction among individuals with BD and should be addressed by mental health providers. Additional research on the mechanisms of, and optimal treatment for, smoking and nicotine dependence in this population is desperately needed.
Cigarette smoking is significantly more prevalent among individuals with bipolar disorder (BD) than among individuals without the disorder (1,2) and has dire consequences. For example, a recent study examining mortality rates among consumers of public mental health services estimated that, on average, individuals with BD and other major mental illnesses die up to three decades earlier than their general population counterparts (3). Although some of these untimely deaths resulted from suicides or accidents, the most frequent causes of death were cancer, cardiovascular, cerebrovascular, and respiratory diseases—conditions for which tobacco use is a known cause of early mortality (4).
Despite the scope of the problem and the implications of tobacco use in BD, limited systematic research is available on this topic. Consequently, the purpose of this review is to synthesize the existing literature and delineate a research agenda, toward a goal of improving knowledge of the underlying mechanisms of, and treatment for, nicotine dependence in individuals with BD. Because smoking is the primary means by which tobacco is consumed and previous research on tobacco use among individuals with BD has focused almost exclusively on smoking, we restricted our review to cigarette smoking. In the sections that follow, we: 1) summarize the literature on prevalence and consequences of cigarette smoking among individuals with BD; 2) examine possible genetic, biological, and psychosocial risk factors that may explain the high prevalence of smoking in this population; 3) enumerate the challenges of smoking cessation and discuss implications for treatment; and 4) identify gaps in the literature as a means of guiding future research.
For this selective review, we conducted a PubMed search of English-language articles using the following search strategy: (bipolar disorder OR mania) AND (tobacco OR nicotine OR smoking). This search yielded 262 hits, the majority of which were not relevant to this review. Of the identified articles that were relevant and that we selected for inclusion, only 13 were empirical studies in which cigarette smoking among individuals with BD was the primary focus (1,2,5–15). Of note, 12 of these 13 studies were published after 2004, suggesting that attention to the relationship between smoking and BD may be increasing.
Estimates of the prevalence rate of current smoking among individuals with BD range from 30% to 70% (1,6,16–19). Although some of this variability can be attributed to small sample sizes and non-representative clinical samples, even population-based prevalence studies have produced disparate findings. For example, data from the 1992–1993 National Comorbidity Survey (NCS) suggested that the prevalence of smoking in BD was 69% (16), whereas the more recent 2007 National Health Interview Survey (NHIS) showed a prevalence rate of 46% (19). Despite this variability in prevalence estimates, controlled population-based (16,19,20) and clinical studies (7–9,21) conducted in the U.S. and Europe have consistently demonstrated that the prevalence of smoking is approximately 2 to 3 times higher among individuals with BD than in the general population.
Epidemiologic data also suggest that the prevalence of smoking in BD is high relative to other Axis I psychiatric disorders. In the NCS study, the prevalence of current smoking in BD (69%) exceeded that of individuals with lifetime nonaffective psychosis (49%), drug abuse or dependence (49%), generalized anxiety disorder (46%), posttruamatic stress disorder (45%), alcohol abuse or dependence (44%), panic attacks and phobias (36–40%), and major depression (37%) (16). However, the reliability of comparisons among the diagnostic groups in the NCS study are questionable due to the relatively small sample size within groups. In the more recent and larger NHIS study, individuals with BD were found to have a significantly lower smoking prevalence rate than individuals with schizophrenia (46% vs. 59%) (19). Although the NHIS data set is not without limitations (i.e., limited number of Axis I disorders assessed, use of self-reported diagnoses), the finding that there is a higher prevalence of smoking among individuals with schizophrenia than among those with BD has also been demonstrated in large clinical samples (17). Nonetheless, it appears that BD is among the Axis I psychiatric disorders with the highest prevalence of cigarette smoking and nicotine dependence.
The high prevalence rates of smoking among individuals with BD appear to be attributable to both increased likelihood of initiating smoking and decreased likelihood of successfully quitting. In terms of initiation, in the only controlled study of tobacco use in adolescents with BD, Wilens and colleagues (10) reported that adolescents with BD are more likely to initiate smoking than age-matched peers without psychiatric disorders (i.e., 22% vs. 4%). Additionally, the lifetime prevalence of smoking among individuals with BD is approximately 80%, which is roughly double that of the general population (16). Cross-sectional and retrospective data from several studies also indicate that individuals with BD are more likely than individuals in the general population to be heavy smokers (7,8) and to be nicotine dependent (20), which can make cessation more challenging. Indeed, although prospective data on smoking cessation are lacking, estimates of the proportion of ever-smoking individuals with BD who successfully quit is lower than the proportion of the general population who quit smoking (with estimates of 8–16% for individuals with BD vs. 33–43% for the general population) (7,16). Of course, these are differences in lifetime quit rates, which are not the same as differences in the likelihood of successful quitting on any given attempt.
Mechanisms underlying the relationship between smoking and BD are likely complex and multifactorial, including genetic and environmental factors as well as their interactions. Research on these mechanisms is sparse in BD, but several decades of research on the relationship between depression and smoking suggest causal pathways in both directions (22,23), and that these relationships may also be explained by common or correlated risk factors (24). In the sections that follow, we outline evidence suggesting that the co-occurrence of smoking and BD may also be attributable to bidirectional influences as well as common risk factors. In doing so, we draw upon the more developed literature on mechanisms underlying the relationship between depression and smoking to supplement the available literature on BD and smoking and to highlight potential avenues for further research.
One potential explanation for the high rate of smoking among individuals with BD is that bipolar symptoms may increase the risk of initiating or maintaining regular smoking. Because of tobacco smoke’s monoamine oxidase (MAO)-inhibiting effects (25) and nicotine’s capacity to stimulate the release of neurotransmitters that improve mood and induce feelings of pleasure (e.g., serotonin and dopamine), the “self-medication hypothesis” has been invoked to explain the greater risk of smoking among individuals with depression. As described by Khantzian (26), this hypothesis suggests that an individual’s preference for, or use of, a specific substance of abuse is motivated by the propensity of that substance to pharmacologically alter an undesired emotional state.
In support of the self-medication hypothesis, depressed smokers are more likely than non-depressed smokers to report smoking as a means of reducing negative affect and increasing positive affect (27). Additionally, nicotine replacement reduces post-cessation negative affect in depressed smokers (28), and experimental evidence indicates that the mood-enhancing effects of nicotine are heightened in depressed smokers as compared to their non-depressed counterparts (29). It is not clear, however, whether these findings generalize to smokers with BD.
In addition to its effects on mood, nicotine also has a modest capacity to enhance cognitive functioning, including attention. In fact, self-medication of attentional deficits has been posited as one of the mechanisms underlying the high rate of smoking among individuals with attention-deficit/hyperactivity disorder (ADHD) (30,31). Given the high rate of co-occurring ADHD among individuals with BD (32) as well as the cognitive deficits that are evident both during and between affective episodes in BD (33), the cognitive enhancement associated with nicotine use may represent another form of self-medication underlying the relationship between smoking and BD. However, in the only study to date that has evaluated the relationship between smoking and cognitive performance in adults with BD (all of whom were euthymic at the time of testing), smokers scored higher on the subjective, but not the objective, assessments of cognitive functioning (11). Thus, although this was a small pilot study, these results suggest that the expectancies regarding cognitive benefits may be more influential in motivating individuals with BD to smoke than the actual effects of tobacco use on cognitive performance.
In addition to BD symptoms, medications used in the treatment of BD, such as certain antipsychotics and antiepileptics, may also enhance the risk of developing regular or dependent smoking. For example, among smokers with schizophrenia, nicotine has been demonstrated to offset some of the side effects associated with antipsychotic medications by improving performance on objective indicators of attention and memory (34). Additionally, treatment with higher dosages of antipsychotics has been linked to greater motivation to smoke for stimulatory effects in smokers with schizophrenia (35). Some medications may also impact smoking by altering the metabolism of nicotine, as suggested in a recent preliminary report indicating that carbamazepine may induce nicotine metabolism through its actions on CYP2A6 (36). Although it has not been conclusively demonstrated that medication-induced changes in nicotine metabolism alter smoking behavior, there is evidence that individuals who are faster metabolizers of nicotine tend to be heavier smokers (37), supporting the potential clinical significance of these findings.
A final consideration regarding the hypothesis that BD increases the risk of smoking is that, by virtue of having BD, individuals may be exposed to environments that are more conducive to smoking. For example, being in treatment settings where mental health professionals either implicitly or explicitly condone smoking and where a high proportion of patients and/or staff smoke are environmental influences that support initiation or continuation of smoking.
Another possible explanation for the high rates of smoking in BD is that smoking may increase the risk of developing BD. For example, smoking may lead to alterations in neurophysiology that unmask an underlying vulnerability to affective episodes. Consistent with this hypothesis, there is evidence from both preclinical and clinical studies to suggest that chronic exposure to nicotine may increase risk of developing depression by desensitizing nicotinic acetylcholine receptors in the brain’s limbic system (38). It has also been proposed that longer-term exposure to nicotine can induce depression through actions on serotonin pathways in the hippocampus, but smokers are protected from these effects as long as they continue to smoke (39).
Although there has been no systematic study of this phenomenon as it relates to onset of BD, it has been suggested that smoking cessation may precipitate an affective episode in some smokers who have already been diagnosed with BD. For example, one study that included a very small subsample (n=8) of smokers with BD demonstrated onset of depression following smoking cessation in 38% (n=3) of those individuals, which exceeded the incidence of emergent depression in smokers without depression histories (40). There are also rare case reports of episodes of mania coinciding with smoking cessation (41,42), and in one small (n=5), placebo-controlled trial of bupropion (Zyban®) for smoking cessation in BD, 40% (n=2, both of whom were taking placebo) of the participants developed hypomania during the treatment period (12). However, the quality of evidence from case reports and small-sample studies such as this one is low, and it would be premature to conclude from these reports that smoking cessation is a precipitant of affective episodes in smokers with BD. Rather, these results suggest that this area warrants further study.
BD and smoking may also be linked through common risk factors, including both genetic and environmental influences. Regarding shared genetic risks, several overlapping candidate genes for BD and smoking have been identified, including genes that encode: 1) catechol-O-methyltransferase (COMT); 2) the dopamine transporter; and 3) the serotonin transporter (13). Moreover, a number of neurochemical systems have been implicated in both BD and nicotine dependence, with most attention focused on dopamine, serotonin and norepinephrine (43–48).
Recent studies in both patients with BD and tobacco smokers have demonstrated structural and functional differences in overlapping brain regions across the two groups. Specifically, the anterior limbic network—a system which is intimately related to functions and traits which are central to BD (e.g. impulsivity and affective regulation in the context of a manic episode) (49) and which are hypothesized to relate to the pathophysiology of nicotine dependence (e.g. dysfunctional reward processing) (50)—has received significant attention. Several prefrontal subregions, including the dorsolateral prefrontal cortex, ventral prefrontal cortex, and anterior cingulate cortex exhibit volumetric differences among adults with BD and among smokers relative to comparison subjects (51–55). In terms of functional differences, functional magnetic resonance imaging (fMRI) studies of patients with BD have demonstrated increased brain activation in important regions of the cortico-limbic pathways responsible for emotional regulation (i.e. prefrontal cortex, anterior cingulate cortex, amygdala, thalamus and striatum) (see reference 49 for recent review). Similarly, fMRI studies of nicotine dependent individuals without BD have observed differences in the functional activation of many of these mesocorticolimbic structures, often in response to fMRI tasks involving smoking cues (56,57).
In addition to these biological explanations for the high rates of smoking among individuals with BD, potential shared environmental risk factors for BD and smoking likely contribute to their co-occurrence. For example, abuse and other forms of childhood adversity have been linked to a greater risk of developing psychiatric disorders as well as daily smoking (58). Additionally, alcohol and illicit drug use have been implicated in both the development of regular smoking and nicotine dependence (59,60) and in the onset of BD (61,62).
In summary, although the evidence in individuals with BD is sparse, there are some indications that BD may be similar to depression in the mechanisms underlying its relationship with smoking. That is, the relationship is likely bidirectional, at least partly attributable to shared risk factors, and includes genetic/biological and environmental risks. Given the substantial gaps in the literature, however, many questions remain. For example, the extent to which individuals with BD might smoke as a means of self-medication and the extent to which tobacco use might influence the development of BD symptoms have yet to be evaluated. Prospective and longitudinal studies are needed to: 1) evaluate the temporal ordering of smoking and BD onset in order to establish directionality of the relationship; 2) determine the extent to which smoking frequency and intensity are affected by mood state and mood state is affected by smoking; and 3) identify biological risk factors for tobacco smoking in patients with BD as well as neural correlates of smoking cessation treatment response. The addition of more detailed assessments of smoking and nicotine dependence to ongoing and new longitudinal studies of adolescents and adults with BD would help to begin answering some of these questions, as would an increased focus on the process of quitting among smokers with BD who attempt cessation.
Cigarette smoking is a leading cause of mortality in the United States and worldwide, and, as noted previously, it contributes to the significant discrepancies in life expectancy between individuals with major mental illnesses such as BD and the general population (3,17). In addition to dying decades earlier, individuals with BD are also more likely to be living with chronic medical illnesses than individuals without the disorder (63). These illnesses are often known to be caused or exacerbated by cigarette smoking and include conditions such as chronic obstructive pulmonary disease, asthma, hypertension, and cardiovascular disease (64,65). In addition to their greater likelihood of having these diseases, individuals with BD also appear to be developing them at substantially younger ages (63). Moreover, high rates of obesity and the metabolic syndrome are also associated with BD (66–68), and the addition of cigarette smoking to these conditions creates a dangerous combination of risk factors for premature death (69).
Polycyclic aromatic hydrocarbons in tobacco smoke increase the metabolism of some psychotropic medications through actions on the cytochrome P450 system, specifically through induction of the CYP1A2 enzyme (70). This enzyme metabolizes several antipsychotic and antidepressant drugs frequently used in BD patients, including olanzapine (71), clozapine (72), haloperidol (70), and fluvoxamine (73). Cigarette smokers may require as much as a 50% increase in dose of these medications in order to gain symptom relief (74). Notably, heavy smoking is not a prerequisite for such dose adjustments, as the pharmacokinetics of clozapine and olanzapine in smokers who consume 7–12 cigarettes per day are statistically indistinguishable from that of smokers who consume over one pack per day (74).
If tobacco use is not considered when establishing a dosing regimen for medications that interact with tobacco smoke, treatment outcomes may be worsened. For example, one recent study found that smokers had worse treatment outcomes than nonsmokers when taking olanzapine or haloperidol, but not when taking divalproex (which is not metabolized by CYP1A2) or placebo (6). Additionally, case report level evidence suggests that a substantive increase in smoking can lead to an increase in psychiatric symptoms among individuals with BD who are taking olanzapine, ostensibly due to the resultant reduction in medication levels (75).
The impact of smoking cessation on individuals taking these medications is also important to consider, as a reduction or discontinuation of smoking may result in substantial increases in plasma levels and an increased likelihood of adverse reactions or toxicity. Several published case reports in this area have included accounts of increased extrapyramidal side effects, seizures, sedation, and worsening of psychiatric symptoms that commenced with smoking cessation and remitted with a reduction in clozapine or olanzapine dosage (75,76). Therefore, clinicians must be vigilant in monitoring adverse effects and adjusting medication levels as needed when patients are trying to quit smoking.
In addition to the pharmacokinetic effects of tobacco smoke on the metabolism of certain psychotropic medications, nicotine can pharmacodynamically impact the efficacy of some medications used in the treatment of psychiatric or physical conditions. For example, nicotine’s activation of the sympathetic nervous system may reduce the sedating effects of benzodiazepines and the capacity of β-blockers to lower heart rate and blood pressure (70). Individuals who use tobacco or nicotine replacement therapies may require higher dosages of these medications, and conversely, individuals who discontinue use of nicotine-containing products may require a dosage reduction.
In summary, the negative impact of tobacco use on the lives of people with BD is far-reaching, yet tobacco’s detrimental effects on both the quantity and the quality of life among individuals with BD are entirely preventable. Nonetheless, there could be several impediments to successful tobacco cessation in this group of smokers. In the sections that follow, we review some of the likely challenges and outline the principles of evidence-based tobacco treatment for tobacco cessation. We also discuss some unique treatment considerations for this special population of smokers, highlighting areas where additional research is needed to guide clinical practice.
Although there are no adequately-sized, prospective tobacco cessation studies in smokers with BD, evidence from epidemiologic studies suggests that quit rates for smokers with BD may be worse than for smokers without the disorder (7,16). These low quit rates could reflect the unique challenges that smokers with BD face when they attempt to quit, which we review in this section (see Table 1 for a summary). The current dearth of research focusing specifically on smokers with BD requires some selective extrapolation from the literature on smoking cessation among individuals with depression and schizophrenia. Because of the uncertain generalizability of these findings to smokers with BD, conclusions derived from them should be considered tentative.
Although the evidence is not unanimous, prior research has demonstrated that smoking is correlated with more severe mood symptoms in BD, including more severe symptoms of depression, rapid cycling, greater overall illness severity, and increased likelihood of suicidal ideation and behavior (1,14,15). The influence of mood state on the success of a smoking cessation attempt among individuals with BD has not been investigated, but previous research suggests that depression is a risk factor for smoking cessation treatment failure (28,77). Although this finding has been disputed (78), it may be specific to multiple major depressive episodes (40) that typify recurrent major depressive disorder or bipolar disorder (1) rather than single-episode major depressive disorder. The link between recurrent depressive episodes and reduced smoking cessation success may be attributable to the increased likelihood of depressive symptoms either preceding or following attempts to quit smoking (77, 79) and the use of smoking as a form of coping with negative affect or mood (27,28,80). In fact, there is strong evidence linking smoking relapse to post-quit negative affect (81), and smokers with a history of depression may be particularly vulnerable to affect-related relapse.
The effects of mania or hypomania on a smoking cessation attempt have not been as systematically evaluated as depression has been, but there is reason to believe that these mood states may also represent a threat to smoking abstinence. For example, the impulsivity that is characteristic of mania and hypomania is a likely barrier to cessation. Indeed, prior research with non-psychiatric smokers has demonstrated a relationship between impulsivity and smoking relapse (82).
Numerous studies have documented strong relationships among cigarette smoking, alcohol and marijuana use, and BD. We previously suggested that recent use of marijuana and alcohol was two to three times more prevalent among bipolar smokers as compared to nonsmokers (2). Additionally, over half (53%) of participants who currently smoked cigarettes met lifetime criteria for a cannabis use disorder (vs. 12% of the nonsmokers), and over one-third (34%) met criteria for an alcohol use disorder (vs. 15% of the nonsmokers) (2). Ostacher and colleagues (15) also found that individuals with BD who had lifetime alcohol and other substance use disorders had 2 to 3 times the odds of being an ever-smoker as compared to those without a substance use disorder history.
Although there have been no prospective studies of the effects of alcohol and drug use on smoking cessation in individuals with BD, studies involving smokers without psychiatric disorders have demonstrated that both alcohol and cannabis use interfere with smoking cessation. Laboratory evidence suggests that alcohol use triggers craving to smoke cigarettes (83), and alcohol consumption has been linked to smoking relapse in a number of smoking cessation studies (84–86). Likewise, cannabis use has been linked to a decreased likelihood of quitting smoking in both epidemiologic (87) and treatment (88) studies. Finally, alcohol and drug use may indirectly threaten the success of smoking cessation efforts by exacerbating psychiatric symptoms (89).
Use of nicotine, the primary addictive substance in tobacco, can lead to a state of dependence similar to other substances of abuse. This includes a withdrawal syndrome marked by symptoms such as irritability, anxiety, dysphoria, impaired concentration, and increased appetite or weight gain. Nicotine dependence and withdrawal are critical factors in the maintenance of smoking, and more severe dependence or withdrawal are associated with more difficulty in quitting smoking (90). Smokers with psychiatric disorders—most notably those with serious mental illness and/or substance use disorders-- tend to present with more severe nicotine dependence and nicotine withdrawal than smokers without these illnesses (91–93). Given that some studies have found that ability to quit smoking is inversely associated with the amount of cigarettes smoked (94), the higher prevalence of heavy, daily smoking among individuals with BD (7) represents an additional challenge to cessation. In a related fashion, one aspect of nicotine withdrawal—the increased appetite and weight gain experienced by the majority of people who quit (95)—may be especially problematic for smokers with BD. Weight concerns are common among smokers who are preparing to quit, especially among women, and these concerns can translate into failure to initiate or sustain a quit attempt (96–98). Because weight gain is also a side effect of several of the medications used to treat BD, additional weight gain subsequent to smoking cessation may not be well-tolerated.
Lack of support for quitting and a high density of smokers in one’s social environment are two common socioenvironmental barriers to successful smoking cessation, and both of these factors are likely to plague smokers with BD. In a preliminary study of a smoking cessation intervention for mental health outpatients, participants reported a perceived lack of support from partners for any behavioral change, including smoking cessation (99). It is likely that a portion of the partners in this study who were perceived as unsupportive were smokers themselves, and the presence of smokers in the household is a relatively robust predictor of a poor smoking cessation outcome (99–103).
Mental health professionals who treat patients with BD are uniquely positioned to offer support and treatment for smoking cessation, yet research indicates that they seldom do so. For example, a study of psychiatrists’ practices with respect to tobacco cessation interventions suggested that providers were fairly conscientious in assessing patients’ smoking status (status identified in 76% of visits), but not in providing cessation counseling (12% of visits) or pharmacotherapy (0% of visits) (104). Similarly, a survey of psychologists indicated that only 29% of respondents asked all patients if they smoke, and only 31% advised all of their patients to quit if they were aware that the patient smoked (105).
Many clinicians report concerns that cessation of smoking will lead to worsening of psychiatric symptoms; however, there is little empirical evidence to support this contention. For example, several studies found that smoking cessation did not lead to a clinically significant deterioration in psychiatric status among individuals with schizophrenia (106,107). With regard to depression, although it is common for recent quitters to experience short-lived dysphoria as part of nicotine withdrawal, onset of a major depressive episode is considerably less common, with estimates of the incidence generally ranging from less than 1% to 7% (77,108,109). Onset of a major depressive episode may occur more frequently in smokers with a history of major depressive disorder (MDD) (108), but not all studies agree on this point (110). As noted previously, although there have been a few published case reports (41,42) and treatment studies involving only a handful of smokers with BD (12,40), the hypothesis that smoking cessation worsens BD symptoms has not been tested and seems unlikely.
At the organizational level, despite the potential of mental health treatment programs to facilitate change in health behaviors, cigarette smoking has historically been ignored or, in some cases, even encouraged in treatment settings. However, these historical trends are changing, and many treatment programs have become tobacco-free and are providing tobacco cessation treatment as part of the clinical programming. Reports of successful transitions to tobacco-free facilities have been accumulating in the literature, and there is evidence that, when planned and implemented thoughtfully, this transition does not adversely affect patients or staff. In fact, some studies have documented improvements in patient behavior and patient-staff interactions as a result of smoke-free policies (111), and a tobacco-free environment may increase patients’ self-efficacy and intention to quit (112). At the same time, a high prevalence of patients who smoke; treatment providers who are reluctant to address smoking, particularly if they are smokers themselves; and the lack of staff training in tobacco treatment remain critical barriers to promoting smoking cessation in some mental health treatment settings (113).
In summary, smokers with BD are likely to experience greater challenges to tobacco cessation than the average smoker, including chronic mood dysregulation, more severe nicotine dependence, and limited support for quitting. However, these challenges can be overcome, and the extent of their impact on the outcome of a cessation attempt in smokers with BD has not yet been examined. Regardless, mental health and/or addictions treatment providers have the expertise to assist smokers with BD in overcoming these challenges, yet research suggests that these individuals may not be consistently addressing clients’ tobacco use. Integration of tobacco treatment into mental health and addictions treatment would therefore expand the current reach of efforts to reduce the prevalence of smoking among people with BD and other psychiatric disorders, and reduce the disproportionate burden of tobacco-related disease borne by these individuals.
Effective interventions for smoking cessation range from brief advice to quit up to combined behavioral and pharmacological treatments. The latter is designed to target both the psychosocial and biological factors that are involved in tobacco addiction and is currently considered to be the gold standard of smoking cessation treatment (90). A range of behavioral therapy and pharmacotherapy approaches are currently available to assist smokers in quitting (see the recently-updated clinical practice guidelines  for a comprehensive summary of these interventions and the evidence that supports their use). Importantly, studies examining the efficacy of these interventions for smokers with psychiatric and substance use disorders demonstrate that these treatments are effective in this special population of smokers (114,115). In this section, we sketch the major principles of pharmacotherapy and behavioral therapy for smoking cessation and discuss treatment considerations for smokers with BD.
Current treatment guidelines (90) suggest that all smokers wishing to quit should be offered pharmacotherapy. There are several pharmacotherapies approved by the FDA as first-line treatments for smoking cessation: nicotine replacement therapy (NRT), bupropion sustained-release, and varenicline (Chantix/Champix®) (see Table 2 for an overview of these agents, their efficacy in smokers without mental illness, and considerations for use in smokers with BD). At present, the safety and efficacy profiles of these smoking cessation aids for individuals with BD are unclear due to the lack of controlled trials that include this group of smokers. Consequently, those who prescribe these agents should keep in mind that there may be unknown risks and/or benefits associated with these medications in populations of smokers who are typically excluded from clinical trials, including smokers with BD.
NRT is available in a variety of formulations (i.e., as over-the-counter lozenge, gum, or patch; and prescription-only inhaler and nasal spray), all of which are of similar efficacy (90). Combination NRT, in which a shorter-acting formulation such as nicotine gum or nasal spray is added onto longer-acting nicotine patch, has demonstrated some improved efficacy beyond single-formulation NRT in controlled clinical trials among smokers without psychiatric disorders (90) and may be beneficial to heavier smokers by providing a greater percentage of nicotine replacement. Similarly, anecdotal clinical reports suggest that extended treatment with NRT (e.g., up to 6 months) may confer some additional benefits for smokers with psychiatric disorders (94), although extended treatment has not been shown to improve long-term smoking cessation outcomes in the general population (90), and this hypothesis has yet to be tested in smokers with BD. Additionally, prior studies have demonstrated that adequate treatment with nicotine gum can delay, but not entirely prevent, postcessation weight gain (69), one of the likely challenges to smoking cessation in this population.
Bupropion, an antidepressant and nicotinic acetylcholine receptor antagonist, can be used either alone or off-label in combination with NRT as a smoking cessation aid. By itself, bupropion is of similar efficacy to single-formulation NRT, with the combination of bupropion and NRT appearing to be more efficacious than bupropion alone (116). Like nicotine gum, bupropion has also been found to delay postcessation weight gain, thereby reducing one of the potential barriers to smoking cessation (69). One complicating factor related to the use of bupropion for smokers with BD include its potential interaction with other psychotropic medications (e.g., carbamazepine, which induces metabolism of bupropion and decreases plasma levels substantially). (117) There is also the possibility of inducing mood destabilization, although this has been observed only in a small proportion of individuals with BD who were taking the medication as an antidepressant (118,119), and there is no strong evidence to suggest that bupropion induces mania when used as a smoking cessation agent in stably treated BD. Additionally, bupropion currently carries an FDA-mandated boxed warning due, in part, to postmarketing reports of increased risk of neuropsychiatric side effects such as agitation, depressed mood, and suicidality. Thus, individuals with BD who are taking bupropion as a smoking cessation aid should be closely monitored for signs of accelerated mood cycling, and this medication should only be used in conjunction with a mood stabilizing agent. Also, because of the increased seizure risk associated with this medication, it should be used cautiously in combination with other psychotropic medications that increase seizure risk, including some antipsychotics (46).
Varenicline, an α4β2 nicotinic receptor partial agonist, is the most recent addition to the arsenal of smoking cessation pharmacotherapies. Evidence to date suggests that varenicline is among the most efficacious of the available smoking cessation medications (90,120,121). However, there have been reports of exacerbation of psychiatric symptoms associated with varenicline use that prompted a recent warning to this effect by the FDA, and several anecdotal reports of worsening of symptoms of BD have appeared in the literature (122–124). Thus, as with bupropion, smokers with BD who are using varenicline should be monitored carefully for worsening of psychiatric symptoms.
In addition to pharmacotherapy, behavioral interventions also improve the probability of quitting smoking, particularly when used in combination with pharmacotherapy (90). Targeted behavioral interventions that are designed to enhance the efficacy of treatment in psychiatric populations have been developed and tested (125–127), but current evidence is insufficient to recommend these approaches as alternatives to standard counseling. Despite the current unavailability of widespread evidence to support their use, pilot work investigating targeted interventions with specialized content have shown some promise for smokers with psychiatric disorders including depression (125,126,128) and substance use disorders (129). For example, cognitive-behavioral therapy (CBT)-based mood management interventions for smoking cessation have been tested in smokers with a history of depression (125,126,128,130) and, although they were not effective in smokers with a history of single-episode depression, they were effective for smokers who had experienced recurrent depressive episodes in the past (125,131). Given that the majority of individuals with bipolar disorder experience recurrent episodes of depression (1), such an intervention may hold promise in this group as well, but this proposition has yet to be tested.
In general, increasing amounts of treatment contact time are associated with increased rates of smoking cessation success (90). Psychosocial interventions of greater intensity may be particularly important as an adjunct to pharmacotherapy for smokers with a history of depression (126,128). Although some individuals with BD may benefit from brief advice and support alone, they should also be encouraged to consider options for more intensive behavioral interventions, particularly if initial attempts at quitting without intensive support are unsuccessful.
The optimal timing of smoking cessation treatment for individuals with substance-related and other psychiatric disorders has been discussed extensively, but there have been few systematic investigations to date that were designed to answer this important question (113,132). Thus, there is little empirical guidance to determine the optimal timing of a smoking cessation attempt for individuals with BD, and clinical judgment is needed to guide decisions about the best time to intervene regarding tobacco cessation. It is our position that periods of relative stability or euthymia are the ideal points at which to encourage tobacco cessation (133). At the same time, it is ill-advised to discourage a smoker who expresses interest quitting from doing so unless there is clear reason to believe that such an action carries a strong risk of untoward effects and should consequently be delayed. Examples of such times might include points at which medication dosages are being titrated or when there has been a recent increase in psychiatric symptoms.
In summary, a number of effective treatments for tobacco cessation are available to aid smokers who want to quit, and there is solid evidence that these treatments can and do help individuals with psychiatric disorders quit smoking. At the same time, additional research examining their safety and efficacy in smokers with BD would provide better guidance on treatment decisions for this group. Because of the current boxed warnings for bupropion and varenicline, smokers with BD who are using these as cessation agents should be monitored for signs of worsening psychiatric status. Pharmacotherapy provides the best outcomes when used in combination with counseling, and, in general, greater amounts of counseling are associated with increasing odds of successful quitting. However, brief advice to quit is effective and should be provided to all smokers, consistent with Clinical Practice Guideline recommendations (90). More research is needed to determine the optimal timing of treatment, intensity of counseling, and efficacy of targeted treatment approaches for smokers with BD.
Much like alcohol and illicit drug use disorders, cigarette smoking is a prevalent and devastating co-occurring condition in individuals with BD. In addition to the challenges faced by all smokers when they try to quit, smokers with BD face additional hurdles related to the core features of the disorder (i.e., dysregulation of mood) as well as its correlates (e.g., co-occurring alcohol and drug use, heavy smoking/nicotine dependence) during a cessation attempt. However, the benefits of tobacco cessation are innumerable, and prior demonstrations that smokers with other mental illnesses such as depression and schizophrenia can quit smoking successfully with the assistance of evidence-based treatments are encouraging. Nonetheless, there is a significant need for research to identify the best approaches to facilitate long-term smoking abstinence for smokers with BD in order to reduce the prevalence and consequences of this dangerous addiction.
Sources of funding: The writing of this manuscript was supported, in part, by grant awards DA026517 (JLH), DA022221 (MPD), AA016372 (MPD), MH077138 (SMS), and by the Department of Veterans Affairs.
Disclosures: Dr. Heffner has provided consultancy services to Pfizer. Dr. Strawn has received research support from the American Academy of Child & Adolescent Psychiatry and from Eli Lilly and Company. Dr. DelBello has served as a consultant or speaker for Eli Lilly, Schering-Plough/Merck, Bristol-Myers Squibb, and the American Academy of Child and Adolescent Psychiatry. In the past year, Dr. Strakowski has received financial support from CME Outfitters, Consensus Medical Communications, Redo Tour (Adamed), Web MD, American Psychiatric Association, and American Academy of Child and Adolescent Psychiatry for speaking. Dr. Anthenelli provides consultancy and advisory services to Pfizer and GlaxoSmithKline. The Tri-State Tobacco and Alcohol Research Center (Drs. Heffner and Anthenelli) has received research support from Lilly, Nabi Biopharmaceuticals, sanofi~aventis and Pfizer. The Division of Bipolar Disorders Research (Drs. DelBello and Strakowski) has also received research support from AstraZeneca, Eli Lilly, Johnson and Johnson, Shire, Janssen (Johnson & Johnson), Pfizer, Bristol Myers Squibb, Repligen, Martek, Somerset, Sumitomo, Nutrition 21, and GlaxoSmithKline.