|Home | About | Journals | Submit | Contact Us | Français|
As mortality due to AIDS-related causes has decreased with the use of antiretroviral therapy, there has been a rise in deaths related to non–AIDS-defining illnesses. Given the exceedingly high prevalence of cigarette smoking among individuals living with HIV infection, tobacco has been implicated as a major contributor to this paradigm shift. Evidence suggests that smoking-related illnesses, such as cardiovascular disease, respiratory illnesses, and certain malignancies, contribute substantially to morbidity and mortality among HIV-infected persons. In this review, we summarize the adverse health consequences of smoking relevant to HIV-infected individuals and discuss smoking cessation in this unique population, including a discussion of barriers to quitting and a review of studies that have examined smoking cessation interventions.
Although the prevalence of cigarette smoking among adults in the United States has declined to approximately 20%, studies have consistently indicated that the prevalence of smoking among persons living with HIV infection remains 2- to 3-fold higher than that of the general population (1–6). For example, in 1993, the prospective Pulmonary Complications of HIV Infection Study reported a smoking prevalence of 54% (5). In 2000, a survey of patients receiving HIV ambulatory care from clinics in New England found more than 70% to be current smokers (2). More recently, a questionnaire-based study of New Yorkers living with HIV reported the smoking rate to be 59% (7). As HIV infection is increasingly transformed into a chronic disease, this high prevalence of cigarette smoking has profound health implications for HIV-infected populations (6, 8–11). For example, cigarette smoking is independently associated with lower scores for general health perception, physical functioning, bodily pain, energy, and cognitive functioning (6, 8–10). Furthermore, cigarette smoking has been identified as a significant cofactor in premature development of HIV/antiretroviral-related cardiovascular disease (12–20).
Before the widespread availability and use of combination antiretroviral therapy (ART), the relevance of cigarette smoking as a contributing factor to the development of pulmonary complications was unclear (21–24). Nevertheless, increasing data support that HIV-infected smokers living in the ART era are unusually susceptible to a number of respiratory complications. These include a heightened risk for lower respiratory tract infections (25–28), chronic obstructive pulmonary disease (COPD), and lung cancer.
In this review we summarize the effects of cigarette smoking on a number of health outcomes relevant to the HIV-infected population and focus on issues related to smoking behavior and smoking cessation.
Studies published in the combination ART era demonstrate a clear link between smoking status and the risk of bacterial pneumonia (11, 25, 26). For example, the incidence of bacterial pneumonia was recently reported in SMART (Strategies for the Management of Antiretroviral Therapy), a multicenter study of 5,472 HIV-infected participants randomized to receive continuous or intermittent antiretroviral therapy (25). In both treatment groups, current smoking was a significant risk factor for the development of bacterial pneumonia; in the continuous treatment group, smoking was the single greatest predictor for the development of bacterial pneumonia (hazard ratio, 3.01) (25). Similarly, in a multicenter prospective study of 885 HIV-infected women, the most prominent clinical risk factor for the development of bacterial pneumonia in the ART era was current smoking (26).
Although studies performed in the precombination ART era failed to demonstrate an association between smoking status and the risk for Pneumocystis carinii pneumonia (PCP), a recent study suggests that smokers in the modern ART era are more likely to develop PCP than nonsmoking individuals with HIV. Miguez-Barbano and colleagues performed a case-control, cross-sectional analysis of 521 consecutive HIV-infected patients admitted to a New York University hospital and found that HIV-infected smokers were three times as likely as HIV-infected nonsmokers to be hospitalized with PCP (29).
Evidence supports that the lungs of HIV-infected individuals are especially susceptible to the damaging effects of cigarette smoke (30–38). Using a combination of high-resolution computed tomography of the chest and physiologic criteria to define emphysema, Diaz and colleagues identified emphysema in 17 of 114 HIV-infected participants, compared with 1 of 44 HIV-uninfected control subjects matched for age and smoking history (P = 0.025) (37). Although these observations were initially made predominantly in the precombination ART era, recent data suggest that the risk for COPD among those living with HIV infection remains increased even with the widespread availability and use of ART (38). The risk of COPD in HIV is reviewed in detail in this issue by Morris and colleagues.
There are numerous reports demonstrating an increased risk of lung cancer among HIV-infected individuals, with adjusted risk associated with HIV infection ranging from 2.0 to 7.0 (39–43). Given the prevalence of cigarette smoking in the HIV population, it has been difficult to determine whether lung cancer risk is related solely to the increased prevalence of cigarette smoking. Studies examining this question are reviewed in detail in this issue by Kirk and colleagues. Other HIV-associated malignancies, such as anal and cervical cancer, have a higher prevalence in HIV-infected smokers compared with nonsmokers (44, 45). This association has been explained by higher levels of human papilloma virus (HPV) variants, such as HPV16 and HPV18, found in smokers (46).
Cardiovascular complications have become a major source of morbidity for HIV-infected individuals living in the ART era (47–53). Although HIV-infected persons have a higher risk of diabetes, hyperlipidema, and hypertension than HIV-negative individuals (50–52), the high prevalence of cigarette smoking is clearly an important contributor to cardiovascular morbidity. In a prospective, observational study by Barbaro and colleagues (54), previously untreated and asymptomatic HIV-infected patients were placed on an ART regimen with a protease inhibitor or without a protease inhibitor. The cumulative annual incidence of cardiovascular-related events was higher in the protease inhibitor group (P < 0.001), with a significantly higher annual incidence of myocardial infarction and metabolic alterations. This acceleration in cardiovascular-related events was greater among men and among those who smoked more than 20 cigarettes per day.
Although protease inhibitors and smoking both contribute to the increased risk of cardiovascular morbidity and mortality, results from the Data Collection on Adverse Events of Anti-HIV Drugs (DAD Study Group) show that current smoking is associated with greater risk of myocardial infarction (relative risk, 2.83; 95% confidence interval [CI], 2.04–3.93) than protease inhibitor therapy (relative risk, 1.16; 95% CI, 1.10–1.23) (47). The increased risk of cardiovascular-related illness also translates into an increased risk of cardiovascular mortality in HIV-infected individuals. De Socio and colleagues (55) compared 403 HIV-infected subjects free from overt cardiovascular disease with 96 control subjects and found that the 10-year estimated risk for cardiovascular mortality was 1.23 ± 2.3% versus 0.83 ± 0.9%, respectively (P = 0.01).
Smoking in the HIV-infected population not only contributes to the comorbid illnesses mentioned previously, but it also has an impact on quality of life and survival. Turner and colleagues (10) studied 585 HIV-infected persons and found that smoking was associated significantly with decreasing scores in all dimensions of health-related quality of life. Using multivariate regression analysis, smoking was significantly associated with decreases in general health perception, physical functioning, bodily pain, energy, role functioning, and cognitive functioning. In the ART era, Crothers and colleagues (56) examined the impact of smoking on quality of life and mortality in subjects enrolled in the Veterans Aging Cohort 3 Site Study. Out of the 867 subjects in the study, 63% were current smokers, and 22% were former smokers. After adjusting for race, age, viral load, CD4 count, and illegal drug and alcohol use, current smoking was associated with a significantly lower quality of life and with a significantly higher mortality compared with never smokers (hazard ratio [HR] 1.99; 95% CI, 1.03–3.86). A subsequent study in the Veterans Aging Cohort 5 Site Study comparing HIV-infected and non–HIV-infected veterans found that smoking was associated with increased comorbid disease and mortality in the HIV-infected group (57).
The recent prospective study by Lifson and colleagues highlights the excess morbidity and mortality attributable to smoking in the HIV-infected population (11). Using data from the SMART trial, these investigators found that, compared with nonsmokers, current smokers had a significantly greater adjusted hazard ratio (HR) for major cardiovascular disease (HR, 2.0), non-AIDS malignancy (HR, 1.8), bacterial pneumonia (HR, 2.3), and all-cause mortality (HR, 2.4). Current smokers also had a significantly greater HR for these complications compared with former smokers, underscoring the potential significance of smoking cessation in this population (11).
Evidence demonstrates that cigarette smoking adversely affects the immunologic response to ART. In a longitudinal study of a large HIV-infected cohort, Feldman and coworkers (58) found that, compared with nonsmokers, smokers receiving ART had poorer viral responses (HR, 0.79; 95% CI, 0.67–0.93), poorer immunologic response (HR, 0.85; 95% CI, 0.73–0.99), greater risk of virologic rebound (HR, 1.39; 95% CI, 1.06–1.69), and more frequent immunologic failure (HR, 1.52; 95% CI, 1.18–1.96).
Given the high prevalence of cigarette smoking and its impact on comorbidities among HIV-infected populations, it is imperative that smoking cessation becomes a priority in the care of HIV-infected individuals. There are a number of barriers and complicating factors that compromise the success of smoking cessation in this unique population. For example, HIV care providers are less likely to identify current smokers and to report less confidence in influencing smoking cessation than non-HIV care providers (59). Managing complications of HIV infection may overshadow smoking cessation discussions and therapies. In addition, HIV-infected individuals may feel that they will ultimately die from HIV, making smoking cessation less of a priority (60). They may also use tobacco to manage HIV-related symptoms and pain (61, 62).
Body image may contribute to fears of smoking cessation in the HIV-infected population because smoking may serve as a means to manage the physical changes accompanying lipodystrophy-associated antiretroviral use (63). Indeed, recent data suggest that HIV-infected smokers may have a greater change in body mass index (BMI) with smoking cessation than would be expected for the general population (64). Among 28 quitters involved in a specialized smoking cessation program at our institution, the increase in mean BMI 3 months after smoking cessation was 2.3 kg/m2 (64), substantially greater than average changes in BMI reported in the general population after smoking cessation (1.1–1.6 kg/m2) (65, 66).
Addiction to drugs and alcohol perpetuates tobacco use (67–69), and the high prevalence of concomitant substance abuse represents a major barrier to smoking cessation in the HIV-infected population (61, 70–72). In a cross-sectional study by Cofrancesco and colleagues (71), the prevalence of illicit drug use was investigated in a cohort of 1,163 HIV-infected individuals compared with 294 control subjects. Eighty-six percent of the HIV-infected participants, versus 67% of the control subjects, reported ever using illicit drugs (P < 0.0001), and 28% of the HIV-infected participants, compared with 16% of the control subjects, reported current illicit drug use (P = 0.0001). Another cross-sectional analysis of 384 HIV-infected individuals found that heavy alcohol drinkers were more likely to smoke (73). Humfleet and colleagues (70) recruited 184 HIV-infected smokers from two outpatient clinics to undergo self-administration of multiple questionnaires measuring their level of nicotine dependence by the Fagerström Test for Nicotine Dependence, desire and readiness to quit, and the characteristics of their social support network. Forty percent of the participants reported alcohol use, 39.7% reported marijuana use, 19.6% reported opiate use, 9.8% reported cocaine use, and 8.7% reported amphetamine use. A large percentage had a history of undergoing treatment for alcohol (33.7%) and for drug abuse (54.9%). The lifetime prevalence of illicit drug use in HIV-infected individuals has been reported to be as high as 84%, with the highest percentage in current and former smokers (61). In a cross-sectional study by Benard and coworkers (74), 509 HIV-infected patients completed self-administered questionnaires about tobacco, alcohol, and illicit drug use as well as nicotine dependence, motivation level, and depression level. Half of the participants were found to be smokers, and there was a significant association between smoking status and excessive alcohol use (P = 0.01) and excessive cannabis consumption (P < 0.001).
Psychiatric disorders also play a strong role in smoking cessation failure. In the general population, over 40% of the cigarettes smoked in the United States are by people with mental illness (75), and comorbid psychiatric disorders have also been associated with lower successful smoking cessation rates (76). This finding has implications for HIV-infected populations because the prevalence of psychiatric disorders in HIV-infected individuals is between 17 and 63% (72, 77–79). HIV-infected smokers have reported higher symptoms of depression and higher consumption of alcohol and illicit drugs than nonsmokers (80). Depression, in particular, has been associated with higher nicotine dependence and lower readiness to quit in HIV (61, 74).
Low socioeconomic status and lack of strong support systems contribute to low success rates in smoking cessation. Socioeconomically disadvantaged people may be less likely to receive smoking cessation advice and treatment (81, 82). Lower education level and lower income and employment status appear to be associated with smoking in HIV-infected individuals (80, 83, 84). In the study by Humfle and colleagues (70), 66.5% of HIV-infected smokers were unemployed, 43.8% had an annual income below $10,000, and 37.4% had unstable housing. Over 40% of the HIV-infected smokers had a social support network comprised predominantly of smokers.
As part of the Lung HIV project, our group has begun to investigate the effects of a specialized smoking cessation intervention on quit rates as well as clinical and biological outcomes in HIV-infected current cigarette smokers. Baseline demographic and psychosocial characteristics of our study population recruited thus far are detailed in Table 1 and demonstrate potential barriers to quitting. For example, nearly 50% have a score of ≥10 on the Beck Depression Inventory, meeting criteria for depression. In addition, a large percentage is unemployed and has low household income.
Despite the numerous barriers that HIV-infected smokers face, a substantial portion exhibits a high motivation to quit (Table 1) (4, 61, 74, 85). Nearly half of HIV-infected smokers in one study was in the contemplation stage of the Transtheoretical Model, and the other half was in the preparation stage (70, 86). This group of HIV-infected smokers had moderate levels of nicotine dependence and had made an average of approximately four prior quit attempts (70). In a study by Mamary and colleagues (4), 72% of smokers in an HIV clinic reported that they had tried to quit previously. At the time of the study, 63% were thinking of quitting, 69% were interested in a group cessation program, and 82% were interested in pharmacotherapy to facilitate smoking cessation. One French study showed that 33% of HIV-infected smokers had attempted to quit at least once, and attempts were greater among older patients, those with longer duration of HIV illness, and those with previous quit attempts (87).
The impact of tobacco cessation programs in HIV-infected smokers has been addressed in several, mostly small studies that have primarily involved nicotine replacement therapy (NRT) (Table 2). A study by Elzi and colleagues (88) showed that HIV-infected smokers who participated in a structured cessation program of NRT and counseling had a quit rate of 50%, compared with 15% in a historical control group. At 12 months, self-reported smoking abstinence was 38% in the intervention group and 7% in the control group. One randomized clinical trial comparing a program of NRT, self-help materials, and phone counseling with a usual care program comprised only of self-help materials and NRT found that HIV-infected smokers in the phone counseling group had abstinence rates of 36.8%, compared with 10.3% in the usual care group (P = 0.0059) (89). Nevertheless, the largest intervention study to date found no difference in biochemically confirmed quit rate between NRT (n = 232) and motivationally enhanced counseling plus NRT (n = 212) (90). Quit rates in both groups were low (~10%) (90).
Limited data are available regarding pharmacologic therapy other than NRT for smoking cessation in HIV-infected populations. Pedrol-Clotet and colleagues found that of 21 HIV-infected smokers on ART and buproprion, 38% remained abstinent for greater than 1 year, and no clinically significant drug interactions were identified (91). A small Spanish study examining 18 HIV-infected smokers on varenicline and ART revealed five cases of nausea and six cases of sleep disturbance; none warranted treatment interruption. The abstinence rates after 3 and 6 months were 6 of 18 and 5 of 18, respectively (92). Although such limited data suggest that HIV-infected individuals can tolerate varenicline, the high prevalence of depression and other psychiatric disorders warrant caution when prescribing this agent to HIV-infected patients because varenicline has been linked to increased risk of suicidal ideation and worsening of psychiatric conditions in individuals with underlying disease.
There are several potential interactions between ART and smoking cessation pharmacotherapy. For example, ritonavir, a potent inducer of human cytochrome P4502B6 (CYP2B6) (93), combined with lopinavir can significantly decrease plasma concentrations of buproprion (94). Conversely, in vitro studies have shown that several protease inhibitors inhibit CYP2B6 (95) and increase the potential for toxic levels of buproprion, which can manifest as seizures. A case series of 10 patients using buproprion in combination with nelfinavir, ritonavir, or efavirenz for a median duration of 8 months did not identify seizure events in any of the patients (96). However, this was a small study, and none of the patients was on high doses of ritonavir. As such, some precaution should be taken when prescribing buproprion to HIV-infected individuals taking protease inhibitors.
The benefits of smoking cessation in the HIV-infected population have been documented. Vidrine and colleagues enrolled 95 HIV-infected individuals in a smoking cessation study consisting of NRT, counseling, and self-help materials. HIV-related symptom burden and health-related quality of life were assessed 3 months after enrollment in the study. The authors found that the number of abstinent days was associated with lower levels of HIV-related symptom burden (97). Smoking cessation, as part of a therapy of cardiovascular prevention program, has also been shown to reduce cardiovascular risk in HIV-infected individuals (98).
The increased risk of bacterial pneumonia in the HIV-infected population appears limited to current and not former smokers. Bernard and colleagues found a 2-fold risk for bacterial pneumonia comparing HIV-infected current smokers with HIV-infected nonsmokers (27). The risk for former smokers, defined as those who have quit for 1 year or more, was nearly identical to that of never smokers. Data from SMART and HIV Epidemiologic Research Study also demonstrate that the association of cigarette smoking with bacterial pneumonia risk is present only in current smokers; both studies found the risk of former smokers to be almost identical to that of nonsmokers (25, 26).
Persons infected with HIV are unusually susceptible to the adverse effects of cigarette smoking. These adverse effects include a heightened risk for lower respiratory tract infections, COPD, and lung cancer. In addition, smoking increases the risk for other malignancies, contributes to the burden of cardiovascular disease, decreases quality of life, and adversely affects the immunologic response to ART. Despite the burden of cigarette smoking in this population and the associated public health implications as this group ages, there has been surprisingly little smoking-related investigation in HIV-infected cohorts. In particular, research is needed to determine effective smoking cessation strategies for this vulnerable population. It is essential that tobacco cessation programs, designed to combat the specific challenges that HIV-infected smokers face, be available, accessible, and effective. It is also essential that health care providers recognize the increased susceptibility of HIV-infected individuals to cigarette smoke and make smoking cessation a priority for their patients.
Supported by NIH grants R01 HL083478 and HL090313.
Author Disclosure: S.R., M.E.W., S.K., N.R., and A.F. do not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. P.D. received lecture fees from Covidien and received institutional grant support from MPEX Pharmaceuticals, Boehringer Ingelheim, and the Batelle Memorial Institute.