Effects on time to lapse of switching menthol smokers to non-menthol cigarettes prior to a cessation attempt: A Pilot Study

Abstract

Introduction:

Menthol smokers (particularly African Americans) have lower cessation success rates than non-menthol smokers. With bans being considered on characterizing menthol flavor in cigarettes, data is needed regarding how switching to non-menthol cigarettes impacts cessation measures.

Methods:

In this randomized pilot study, African American menthol cigarette smokers interested in quitting smoking either continued smoking menthol cigarettes (n=60) or switched to non-menthol cigarettes (n=62) for a one month period prior to a cessation attempt. The primary endpoint was time to smoking lapse (i.e., time from quitting until any smoking). Additional endpoints included time to smoking relapse (i.e., number of days from quitting until the first of seven consecutive smoking days) and difference between groups in subjective measures.

Results:

After attempting to quit, the non-menthol cigarette group had indications of delayed time to lapse (HR: 0.82; 95% CI: 0.55, 1.22; p=0.33) and time to relapse (HR: 0.67; 95% CI: 0.42, 1.06; p=0.09) although these were not statistically significant. Post-hoc analyses suggest that observed differences were largely due to a smaller proportion of participants in the non-menthol group relapsing within the first day of quitting (21% vs. 40%; p=0.05). Values of other measures assessed post-cessation were largely similar between groups.

Conclusions:

These data suggest that among African American smokers, a menthol cigarette ban would not undermine short-term cessation measures and may result in some benefits. Future research is needed to assess longer term cessation rates and to identify interventions to maximize cessation success in the event of a menthol ban.

INTRODUCTION

The passage of the Family Smoking Prevention and Tobacco Control Act resulted in a ban in the United States of all characterizing flavors in cigarettes except for menthol and tobacco.¹ In March 2018, the Food and Drug Administration issued an advanced notice of proposed rulemaking calling for information regarding the role of flavors (including menthol) in the initiation, use, and cessation of tobacco products suggesting that a ban on characterizing menthol flavor is being considered.² Regulatory action targeting menthol would provide an optimal time for interventions that increase cessation.

One concern regarding characterizing menthol flavor in cigarettes is that it impedes successful smoking cessation, particularly among African Americans. Menthol cigarettes are smoked by over 80% of African American smokers³ and a meta-analysis found lower cessation rates among black / African American menthol vs. non-menthol cigarette smokers with no significant effect observed in white smokers.⁴ Qualitative reviews similarly found that menthol smokers, particularly African Americans, are less likely to successfully quit than are non-menthol smokers.^5–8 Recruitment in the current study was therefore limited to African Americans.

Initial data from a cohort study conducted in Ontario, Canada where menthol was banned suggest increased cessation attempts and cessation rates among menthol smokers.⁹ Studies in which smokers switched from menthol to non-menthol cigarettes have reported fewer cigarettes smoked, decreased craving severity and increased motivation to quit.^10,11 However, the effects of switching to non-menthol cigarettes on subsequent cessation success are not known.

This pilot study assessed if switching to non-menthol cigarettes is an effective first step to cessation for menthol smokers interested in quitting. We hypothesized that switching to non-menthol cigarettes prior to a cessation attempt delays the time to smoking lapse. Time to smoking lapse was selected as the outcome measure in this pilot study since it is a continuous variable, has been reported to be associated with longer term outcomes, and has been commonly used as an outcome measure in studies.^12–16

METHODS

Participants

Participants were self-identified African Americans, ages 18 to 64 who smoked on average ≥ 5 cigarettes daily and who at study entry rated their motivation to quit smoking “at this time” as ≥ 7 (out of 10). Individuals were excluded who reported using tobacco products other than menthol cigarettes more than 20% of the time; who reported serious, unstable medical or psychiatric conditions; who were taking medications that could impact outcome measures or be affected by changes in smoking status; or women who were pregnant or breast feeding.

Procedures

This study was approved by the University of Minnesota Institutional Review Board. Study visits occurred between October 2014 and July 2018. At a screening visit, written informed consent was obtained. The study’s purpose was described as seeing if switching to non-menthol cigarettes would make quitting easier. Eligible participants scheduled for the baseline visit were randomized to either continue smoking menthol cigarettes or switch to non-menthol cigarettes for a one month period prior to a chosen smoking quit date. At the baseline visit, research staff informed participants of their randomization and provided free cigarettes consistent with their assignment (i.e., usual brand of menthol or preferred brand of non-menthol cigarettes). Participants received approximately 120% of their average daily reported use up to a maximum of about 1.5 packs per day. Since cigarettes were provided in full packs, these percentages were approximations. All participants were encouraged to call the smoking quitline prior to their quit date to obtain support for their cessation attempt with no additional instructions provided regarding how to change their smoking behavior.

Two additional visits occurred before the quit date with a visit also occurring on either the quit date or the day before. Post-quit visits occurred at weeks 1, 2, 4, 6, 8, 12 and 26. Participants were compensated for each visit that they completed. At post-quit visits, no tobacco products were provided and participants were instructed to smoke their assigned cigarette type if lapses occurred. At post-quit visits, participants were asked about lapses and if one occurred, the date and time of its occurrence was determined. Exhaled CO was measured and questionnaires were completed assessing craving and withdrawal symptoms,^17,18 smoking urges,¹⁹ perceived health risk due to the tobacco product being used^20,21 and support for a menthol ban (on a 10 point scale). Total urinary cotinine concentration²² was measured at weeks 0, 1, 2, 4, 8, 12, and 26.

Outcome Measures

The primary outcome is time to lapse (i.e., time from the quit attempt until any smoking). Secondary outcomes include time to relapse (i.e., days from the quit attempt until the first of seven consecutive smoking days),^13,23 and difference between groups in CO confirmed abstinence rates at each visit (i.e., no reported smoking in the past week and exhaled CO < 8 ppm), in exhaled CO and urinary cotinine concentrations and on questionnaire scores.

Statistical Analysis

This pilot study was designed to have approximately 90% power to detect a hazard ratio of ≥ 2.0 (alternatively, ≤ 0.5) between groups if 120 participants were randomized and 15% attrition occurred. Randomization assignments were generated by computerized pseudo-random number generation using permuted blocks of 4 or 6 and stored in a REDCap²⁴ database.

Participants attending the baseline visit were included in the intent-to-treat analysis. Times to lapse and relapse were analyzed using Kaplan-Meier estimates and Cox regression (with Efron’s method for tied event times) starting at the hour of the quit attempt. Follow-up was censored at a participant’s last visit if no lapse was reported. The median event time (not a pre-specified outcome) was compared.²⁵

Other secondary outcomes were analyzed using correlated data models that included multiple observations per subject (from post-quit visits). Fixed effects were included for randomization group, visit, and baseline measurement. Outcomes were modeled using generalized linear mixed models with a random subject intercept. Exhaled CO and cotinine were log-transformed.

Analyses were conducted using R software, version 3.4,²⁶ primarily using the survival,²⁷ geepack,²⁸ and lme4²⁹ code packages.

RESULTS

Of those who completed the baseline visit, 60 were randomized to continue smoking menthol and 62 to switch to non-menthol cigarettes. Of these, 107 completed the week 0 visit, 95 the week 12 visit, and 79 completed the entire study (Supplementary Figure). At enrollment, those randomized to menthol were on average (SD) 47(10) years old, smoked 13(8) cigarettes per day and reported being daily smokers for 23(12) years. Those randomized to non-menthol were 45(11) years old, smoked 12(7) cigarettes per day, and reported being daily smokers for 21(11) years. In both groups, 37% were female, the mean (SD) score on the Fagerstrom Test for Nicotine Dependence³⁰ was 5(2), and the mean (SD) motivation to quit at screening was 8(2).

The hazard ratio for time to lapse was 0.82 (non-menthol vs. menthol) (95% Cl: 0.55, 1.22; p=0.33) (Figure). Median time to lapse in the non-menthol group was 2.55 days vs. 1.07 days in the menthol group (p=0.08, post-hoc test). The hazard ratio for time to relapse was 0.67 (95% CI: 0.42, 1.06; p=0.09) for non-menthol vs. menthol (Figure) with the difference primarily due to lower rates of early relapse in the non-menthol group since 21% of the non-menthol group relapsed within the first day vs. 40% of the menthol group (p=0.05, post-hoc test). Median time to relapse was 42 days in the non-menthol vs. 7 days in the menthol group (p=0.26). In the non-menthol group, time to relapse was similar among those who after a lapse continued to smoke non-menthol cigarettes vs. those who reverted to menthol cigarettes (HR 1.14; 95% CI 0.56 to 2.34; p=0.71), with approximately half reverting to menthol cigarettes.

Figure.

Kaplan-Meier curves of time to first lapse (panel A) and time to relapse (panel B) to smoking.

Point prevalence cessation rates were similar between groups (Supplementary Table 1) as were biochemical biomarkers of smoking (i.e., exhaled CO, urinary cotinine concentrations), craving and withdrawal symptom severity, perceived health risks and support for a menthol ban (Supplementary Table 2).

DISCUSSION

In surveys of menthol smokers, many report that they would quit smoking were menthol cigarettes banned,^31–33 however many smokers may instead switch to non-menthol cigarettes¹⁰ which may be seen by some menthol smokers as a step towards cessation.³⁴ Our randomized study provides preliminary data suggesting that switching to non-menthol cigarettes may have positive effects on short-term cessation measures largely by decreasing the proportion of relapses occurring within the first day of quitting. If confirmed in larger studies, this would suggest that more intensive interventions may be needed during the initial quit period to sustain early success among those switching to non-menthol cigarettes since in the current study smokers were provided with minimal smoking cessation assistance.

As a pilot study, the relatively small sample size precludes drawing definitive conclusions. Additionally, since non-menthol cigarettes were provided, it is possible that changes in cessation success following an actual menthol ban would be different than in this study. It is therefore important to study how characterizing menthol in other tobacco products might influence overall tobacco use patterns were a ban to include only some categories of tobacco products. As cigarette type could not be blinded and participants knew the purpose of the study, it is possible that the expectation of it being easier to quit in those randomized to non-menthol affected the results. Additionally, it is possible that there were differences in nicotine between usual brand menthol cigarettes and the cigarettes chosen by participants randomized to switch to non-menthol and that this impacted the results.³⁵ An additional limitation is that data collection regarding quitline use was incomplete. Since this study only recruited African American menthol smokers who were motivated to quit smoking, it is not known if these results are generalizable to other groups nor is it known if switching to non-menthol cigarettes for a period other than four weeks would affect the results.

CONCLUSIONS

This pilot study suggests that a ban on menthol is not likely to undermine short-term cessation measures among African American menthol smokers interested in quitting and may result in benefit. These data, in addition to being relevant regarding the potential effects of a menthol ban, provide a potential strategy by which current menthol smokers may attempt to quit. Larger studies are needed to confirm the preliminary conclusions reached.

What This Paper Adds.

What is already known on this subject

• Studies indicate that among African Americans, those who smoke menthol cigarettes are less likely to successfully quit smoking than those who smoke non-menthol cigarettes

• Previous studies suggest that in the event of a menthol ban some menthol smokers may attempt to quit but others will likely switch to non-menthol cigarettes

What this paper adds

• We examined, in a randomized study, the effects of African American menthol cigarette smokers switching to non-menthol cigarettes prior to a cessation attempt

• The results of this pilot study suggest that a ban on menthol in cigarettes would not undermine short term cessation measures among current African American menthol smokers and may result in benefits.