Tobacco use remains the leading preventable cause of morbidity and mortality for all racial and ethnic groups in the United States (1). Indeed, smoking is responsible for greater than 30% of all cancer deaths and is a major contributor to heart, lung, and cerebrovascular disease (1). More than 50% of all African American smokers use 10 or fewer cigarettes per day (CPD) and thereby can be described as light smokers (2,3). Although African Americans smoke fewer CPD than the general population, they experience disproportionately greater tobacco-related disease burden, including the highest rates of cancer incidence and mortality (1,4). Unfortunately, African Americans are less likely to receive tobacco use treatment or to achieve abstinence when making a quit attempt, making treatment of this group a public health priority (5,6).

African Americans have been underrepresented in smoking cessation research (7–9), and only two studies have evaluated tobacco use treatment for African American light smokers (9,10). The 2008 Clinical Practice Guidelines for Treating Tobacco Use and Dependence recommend counseling and pharmacotherapy for the treatment of smokers, and call for treatment research with ethnic minority smokers and light smokers, because of limited research with these groups (11). Because most smoking cessation trials have been conducted with predominantly white (of European descent) moderate to heavy smokers, assumptions cannot be made about the benefit of pharmacotherapy for racial and ethnic minority smokers or for light smokers. In addition to smoking fewer CPD, African American smokers are more likely to smoke high-tar and mentholated cigarettes (12,13), to have slower rates of nicotine metabolism (14), and to show higher levels of cotinine (the primary metabolite of nicotine) per cigarette smoked (14–17). Such differences may influence treatment response.

This study builds upon a series of Kick It at Swope (KIS) studies focused on enhancing treatment for tobacco use in African American moderate to heavy smokers (KIS-I study) (8) and African American light smokers (KIS-II study) (9). As part of the KIS-II study, Ahluwalia et al. (9) evaluated the efficacy of nicotine gum vs placebo combined with either health education (HE) counseling or motivational interviewing. The findings demonstrated the efficacy of HE counseling in doubling smoking abstinence relative to motivational interviewing but found no measurable benefit of nicotine gum (9). The sample of light smokers demonstrated variations in daily smoking patterns and a wide range of baseline cotinine levels, suggesting that dosing using nicotine replacement therapy might be challenging. Because sustained release bupropion (bupropion SR) is a non-nicotine medication shown to be effective in producing abstinence in African American moderate to heavy smokers (≥10 CPD) (8), consideration of bupropion for light smokers was warranted. Bupropion is an effective first-line medication for tobacco use treatment, shown to approximately double the abstinence rates at 6 months compared with placebo (8,11,18). Based on previous research, bupropion treatment was expected to aid withdrawal reduction among individuals using the medication to assist quitting (19). The goal of the current Kick It at Swope III (KIS-III) study was to evaluate the efficacy of bupropion in combination with HE counseling for smoking cessation in African American light smokers.

Salivary cotinine–verified and self-reported smoking abstinence rates at weeks 3, 7, and 26 are presented in Table 2. At the primary endpoint of the study, which was cotinine-verified 7-day point prevalence abstinence at week 26, no statistically significant difference between the bupropion SR and placebo groups was observed. Imputing those lost to follow-up as smokers, the difference in long-term smoking abstinence rates at week 26 was not statistically significant between the bupropion SR and placebo groups (13.3% vs 10.0%, OR = 1.39, 95% CI = 0.82 to 2.35, P = .23). However, the initial smoking abstinence rate at week 3 was statistically significantly higher in the bupropion SR group compared with the placebo group (21.5% vs 9.6%, OR = 2.58, 95% CI = 1.56 to 4.22, P < .001). Similarly, at the end of the 7-week medication phase of the intervention, the smoking abstinence rate was statistically significantly higher in the bupropion SR group compared with the placebo group (23.7% vs 9.6%, OR = 2.92, 95% CI = 1.78 to 4.77, P < .001).

Completion of each study session did not differ between the treatment groups (retention rates shown in Figure 1). Retention of participants at week 26 was 71% (192 of 270 participants) in the bupropion SR group and 69% (187 of 270 participants) in the placebo group. No statistically significant difference in number of counseling sessions attended (range 1–6) was observed between the bupropion SR and placebo groups (mean [SD] = 4.9 [1.5] vs 4.8 [1.7], P = .24). Similarly, no difference in self-reported medication adherence at week 3 was observed based on a 3-day pill recall between bupropion SR and placebo groups (mean [SD] = 4.6 [2.1] vs 4.6 [2.3], P = .86). As shown in Table 3, the prevalence of AEs at week 3 was not different between the bupropion SR and placebo groups (P > .05 for all grades of AEs). Furthermore, an analysis of changes in symptoms of depression (Figure 2) and withdrawal (Figure 3) during the study period relative to the initial baseline self-report showed that treatment condition had no effect on changes in the level of depression or withdrawal.

To date, KIS-III is among the first studies to evaluate pharmacotherapy for light smokers (9,10,39). It is the second randomized placebo-controlled study of smoking cessation treatment for African American light smokers (9). Consistent with our previous KIS trials of African American smokers (8,9), study participants were largely low income and predominantly menthol smokers who were highly interested in stopping smoking.

In contrast to the current negative findings of long-term treatment effects, previous placebo-controlled clinical trials of bupropion have demonstrated efficacy in producing long-term abstinence in moderate to heavy smokers (≥10 CPD), with smokers using active bupropion being more than twice as likely to be abstinent at 6 months following treatment compared with those using placebo (11,18). The majority of previous bupropion trials had predominantly white participants. Like these other studies, our KIS-I placebo-controlled trial of bupropion SR for African American moderate to heavy smokers (≥10 CPD) found a statistically significant treatment effect of bupropion SR at the end of the medication phase of treatment (36.0% and 19.0% verified abstinence in bupropion SR and placebo groups, respectively) in addition to sustained abstinence at week 26 follow-up (21.0% and 13.7% in bupropion SR and placebo groups, respectively) (8). A critical question is why these long-term treatment effects seen in heavier African American smokers were not observed in this study of African American light smokers.

It is noteworthy that the current findings demonstrate initial benefit of bupropion SR for light smokers during the medication phase of the study as indicated by the higher abstinence rates in the bupropion SR group at weeks 3 and 7. African American light smokers who were given bupropion SR were almost three times as likely to quit smoking at week 7 compared with those who were given placebo (OR = 2.92, 95% CI = 1.78 to 4.77). This benefit of bupropion SR was seen early on in treatment, as evidenced by differences in abstinence rates between the study groups at week 3. The difference in abstinence rates between bupropion SR and placebo groups during the treatment phase was not related to retention or completion of counseling as no differences in these variables were seen between groups. Furthermore, although some studies have identified an impact of bupropion on reducing withdrawal and symptoms of depression (19), no such treatment effects were evident in this study. This finding is of particular interest given almost a third of our sample reported elevated symptoms of depression. While abstinence rates within the placebo group remained stable between the end of treatment at week 7 and follow-up at week 26, abstinence rates within the bupropion SR group decreased from 23.7% at week 7 to 13.3% at week 26. Notably, within the bupropion SR group, individuals who achieved long-term abstinence demonstrated higher medication adherence early in treatment, reflected in the evaluation of total bupropion levels in the blood at week 3 compared with those still smoking at week 26, suggesting a clinical benefit of bupropion use in facilitating abstinence for some light smokers. Indeed, attention to medication adherence in enhancing pharmacotherapy efficacy merits further consideration.

Limited study of smoking cessation treatment among light smokers has produced mixed findings regarding the benefit of pharmacotherapy for light smokers (9,10,39). Within the KIS-II placebo-controlled study of nicotine gum for African American light smokers (1–10 CPD), Ahluwalia et al. (9) found no statistically significant treatment effect at month 6, with cotinine-verified abstinence of 14.2% and 11.1% for nicotine and placebo groups, respectively. Lack of medication effect may have been related, in part, to challenges in underdosing or underadherence (9). In contrast, within a largely white sample of smokers who used 1–15 CPD and were randomized to placebo or nicotine lozenge, Shiffman (39) found that nicotine lozenge statistically significantly increased carbon monoxide–verified abstinence rates relative to placebo 6 weeks into treatment (45.7% vs 31.1%) and at 12-month follow-up (19.2% vs 10.0%). Gariti et al. (10) compared nicotine patch with bupropion within a racially mixed sample of light smokers (6–15 CPD; 68% African Americans) and found that within the nicotine patch group, quit rates appeared relatively stable over time, with verified abstinence rates of 26.8%, 26.0%, and 23.0% at weeks 12, 26, and 52, respectively. In contrast, bupropion produced abstinence rates of 27.1%, 18.0%, and 15.9% at weeks 12, 26, and 52, respectively, demonstrating higher early abstinence but subsequent decrease in abstinence with time (10). Across treatment groups, African Americans were less likely than whites to be abstinent at the end of the study (10). Benefit of pharmacotherapy and identification of mechanism of action for light smokers warrant additional study, with attention to enhancing initial treatment efficacy and increasing long-term abstinence for African American light smokers.

In this study, although motivation to stop smoking was high within the sample of light smokers, abstinence rates were modest. Indeed, nine of 10 light smokers who received placebo were smoking within the first 3 weeks of treatment, demonstrating that even motivated light smokers had difficulty with initial quit attempts or struggled with early relapse. These findings further support the need to identify effective treatment approaches for light smokers (11). Although this cohort of African American light smokers reported use of approximately 8.0 CPD, findings show notably high levels of cotinine, similar to those seen in our previous study of African American light smokers (9). The majority of these light smokers reported smoking soon after wakening, suggesting physical nicotine dependence (40). Given notable nicotine intake, nicotine dependence, and the early impact of bupropion SR on facilitating abstinence during the medication phase of treatment, these findings support the idea that some light smokers benefit from pharmacotherapy to aid initial cessation. Future studies with African American light smokers could examine extended use of bupropion as a means of building on the initial medication effect to support sustained abstinence over time. Although extended use of bupropion has not been established as a method of relapse prevention (41), a number of studies suggest that long-term use of bupropion (eg, up to 1 year) may promote sustained abstinence (42–44).

Successful recruitment and retention within this study support the feasibility of enrolling African American light smokers into a clinical treatment trial involving biological data collection (7,20,45). The present findings further support the importance of biochemical verification of smoking status, previously identified in KIS clinical trials (8,9), consistent with guidelines on biochemical verification (34).

This study has a few limitations. Generalizability of the findings to other smokers may be limited based on study inclusion criteria that address smoking cessation in African American light smokers motivated to stop smoking and based on study exclusion criteria that largely focus on medical eligibility to use bupropion (20). Although this study design included assessment points standard for smoking cessation treatment studies, that is, end of treatment and 6-month follow-up, the lack of assessment between these points limited the ability to characterize the process of relapse among smokers in the bupropion SR group who experienced initial abstinence. For example, it is not clear if these light smokers more commonly experienced rapid or gradual relapse following the medication phase. Interest in more comprehensive evaluation of smoking behavior change in future study will need to be weighed against considerations of participant burden and study retention.