The Scientific Case for Regulating Menthol: Next to Nothing

Ever since the FDA started regulating tobacco in 2009, a principal objective of tobacco opponents has been a menthol cigarette ban.  Congress could have banned menthol in its authorizing legislation, but Matthew Myers, president of Tobacco-Free Kids and principal architect of the legislation, knew that menthol was a congressional poison pill.  Instead, Congress directed the FDA to conduct a special investigation that prohibitionists hoped would lead to a ban.

The FDA is obligated to act in an unbiased fashion and base its actions on credible scientific evidence.  When, in 2011, the FDA Tobacco Products Scientific Advisory Committee found, on weak evidence, that menthol had a negative impact on public health, a federal judge barred the agency from acting because the committee was tainted by conflicts of interest.

In 2013, the FDA tackled menthol again, issuing a “Preliminary Scientific Evaluation of the Possible Public Health Effects of Menthol Versus Nonmenthol Cigarettes.”  As I pointed out, the agency’s case against menthol was entirely unconvincing. A subsequent filing by Lorillard underscored that fact.    

Late last year, investigators from Reynolds American and Environ International published a series of articles on the impact of menthol cigarettes on current smoking patterns, smoking initiation, measures of dependence and smoking cessation.  While previous reports from anti-tobacco activists used cherry-picked data from various sources, the Reynolds-Environ studies utilized well-documented methods to analyze data from four major surveys: the National Health Interview Survey (NHIS), the National Survey on Drug Use and Health (NSDUH), the National Health and Nutrition Examination Survey (NHANES) and the Tobacco Use Supplement to the Current Population Survey (TUS-CPS).  Following are topline results from the Reynolds-Environ studies, published in Regulatory Toxicology and Pharmacology.

Current Smoking Patterns

Geoffrey M. Curtin et al.  Regulatory Toxicology and Pharmacology Volume 70: Pages 189-196, 2014

“In summary, the current analyses examine menthol cigarette use, overall and within demographic strata, among four nationally representative samples of U.S. adults, and in some instances among youth. Findings from these analyses confirm that adult female smokers are more likely to use menthol cigarettes compared to male smokers, and that a majority of adult non-Hispanic Black smokers use menthol cigarettes; these patterns persist regardless of smoking frequency. Smokers in other race/ethnicity groups may be more likely to use menthol cigarettes compared to non-Hispanic White smokers, but are not more likely to use menthol cigarettes compared to smokers overall. Finally, higher proportions of younger adult smokers (aged 18–25 years) report using menthol cigarettes compared to older adult smokers (aged 26–29 years and 30+ years), but statistically significant differences are inconsistent and numerical differences are small in magnitude; the limited data for youth smokers do not provide consistent patterns of menthol cigarette use.”

Smoking Initiation

Geoffrey M. Curtin et al.  Regulatory Toxicology and Pharmacology Volume 70: Pages 446-456, 2014

“Results from the analyses provided herein indicate that menthol cigarette use is not associated with an earlier age of initiating smoking or a greater likelihood of being a daily versus non-daily smoker; and, that menthol cigarette preference is not greater among new, less-experienced compared to more-established youth smokers.”

Measures of Dependence

Geoffrey M. Curtin et al.  Regulatory Toxicology and Pharmacology Volume 69: Pages 451-466, 2014.

“In summary, findings from the current analyses that examine number of cigarettes smoked per day, time to first cigarette after waking and HSI [the Heaviness of Smoking Index], coupled with evidence from previous studies based on nationally representative populations do not support an evidence-based conclusion that menthol in cigarettes increases dependence among U.S. smokers. Results from the current analyses indicate that menthol smokers do not report smoking a greater number of cigarettes per day compared to non-menthol smokers; that the distribution of menthol compared to non-menthol smokers into categories of time to first cigarette after waking is nearly identical; and, that menthol smokers are generally more likely to be in a lower versus higher HSI (i.e., dependence) category.”

Smoking Cessation

Sandra I. Sulsky et al.  Regulatory Toxicology and Pharmacology Volume 70: Pages 231-241, 2014.

“Data from the NHIS (2005 and 2010) generally suggest no statistically significant association between menthol cigarette use and being a former versus current smoker for any of the three race/ethnic groups, while data from TUS-CPS (2010/11) generally suggest a statistically significant inverse association between menthol cigarette use and having quit smoking, particularly among non-Hispanic Blacks…[The analyses] provided inconsistent results with regard to menthol cigarette use and quitting, both within surveys (i.e., comparing race/ethnicity groups) and between surveys (i.e., same race/ethnicity group across surveys).”

In summary, these comprehensive analyses document that menthol cigarettes are more popular among women, blacks and younger smokers.  The surveys show that menthol plays virtually no role in smoking initiation and dependence.  With respect to whether menthol affects cessation, the NHIS and the TUS-CPS reveal inconsistent results.

Earlier studies of menthol cigarettes showed that they may be associated with lower risk for lung cancer (here) and stroke (here), although this does not mean that they are safer than nonmenthol cigarettes.  The Reynolds-Environ studies provide evidence that there are no significant differences between menthol and regular cigarettes with respect to smoking initiation, addiction to nicotine, or cessation. 

This set of publications use publicly available data; the methods are completely transparent and comprehensive; the findings are broad and highly credible.   

There is no justification for FDA action to ban or otherwise restrict menthol in cigarettes.

Do E-Cigarette Ads Promote Vaping?

The National Cancer Institute is wasting taxpayer dollars on slanted e-cigarette research that didn’t ask or answer an obvious and important question.

Consider the recent NCI-funded study by Drs. Erin Maloney and Joseph Cappella at the UPenn Annenberg School of Communication (abstract here). 

Maloney and Cappella recruited daily smokers, intermittent smokers and former smokers (there were no significant results in the middle group so I won’t discuss them).  They divided smokers into three subgroups: controls who didn’t see an ad, those who saw e-cig ads with vaping (called a cue) and those who viewed ads with no vaping.  Participants answered questions about their inclination to smoke a cigarette, to quit smoking or to continue to abstain.  Maloney and Cappella developed a scoring system to measure responses.  The results they pitched to the media are in the table.

All groups had lower urges to smoke after the experiment, but smokers who saw the cue had less of a lower urge, which was significant in the authors’ scoring.  Similarly, all former smokers had high scores for continuing to abstain, but those who saw the cue had a lower high score. 

E-Cigarette Ads and Urges in Daily Smokers and Former Smokers
	Daily Smokers			Former Smokers
	Cue	No Cue	No Ad	Cue	No Cue	No Ad
Urge to Smoke
Pre-test	3.99	3.83	3.84	1.86	1.69	1.76
Post-test	3.63*	3.14	3.25	1.42	1.34	1.32
Intention to:
Quit	2.14	2.15	2.06	---	---	---
Abstain	---	---	---	12.39*	13.10	13.14

*Significantly different than No Cue or No Ad.

The authors acknowledged that “effect sizes reported in this manuscript were not large.”  In fact, the differences are so small that they may not be meaningful for actual behavior.  Take 12.39 versus 13.14 in the table as an example.  The authors reported that higher numbers are better, and both numbers look “high” when compared with a previous study by Cappella that used the score (abstract here).  He showed former smokers anti-smoking ads in that study and got scores around 3.0 to 3.5.  This looks like e-cigarette ads are far better for former smokers than anti-smoking ads.    

There is a glaring defect in the report.  The researchers collected a lot of basic information (e.g., education, quitting history and time since last cigarette) that could affect how participants responded to questions, but the results weren't adjusted for these important characteristics.  It is possible that the cue, no-cue and no-ad groups had important differences in basic information that affected their scores.

The study’s biggest weakness is that no data was collected on urges and intentions to VAPE.  After all, that is the most important goal of e-cigarette ads, and it is an obvious outcome to measure. 

This study is a failure to communicate.

FDA Preliminary Scientific Evaluation of Menthol: Inadequate for Regulatory Action

The FDA Tobacco Products Scientific Advisory Committee in 2011 reported that menthol had a negative impact on public health.  This week, the FDA released a “Preliminary Scientific Evaluation of the Possible Public Health Effects of Menthol Versus Nonmenthol Cigarettes,” (available here) and announced a 60-day public comment period.  This is the opening salvo in FDA regulatory action that could eliminate or restrict around 30 percent of the American cigarette market (here). 

This post examines the evidence for regulating menthol provided in the preliminary scientific evaluation.

The FDA considered the effects of menthol on several smoking parameters, including smoke chemistry and toxicity, biomarkers, smoking patterns, marketing and consumer perception, smoking initiation and progression, nicotine dependence, and smoking cessation. 

Here are the report’s major findings (I have provided PubMed links to some of the important studies cited by the report):

  

Smoke Chemistry/Toxicity

There were no significant findings: “…menthol in cigarettes is not associated with increased or decreased smoke toxicity.”

Biomarkers

There were no significant findings: “…menthol in cigarettes is likely not associated with increased or decreased levels of biomarkers of exposure.”

Smoking Patterns

The evaluation confirmed what is already widely acknowledged – that menthol cigarettes are more popular among certain groups, especially African Americans, women and, in some studies, younger smokers.

Marketing and Consumer Perception

There were no significant findings: “…a clear relationship cannot be drawn.”

Smoking Initiation and Progression

While the report concluded that “menthol in cigarettes is likely associated with increased initiation and progression to regular use of cigarette smoking,”  this finding is based on remarkably mixed data, described in this passage:

“…There is no indication that menthol smokers first experience cigarette smoking any earlier or later than nonmenthol smokers (Pletcher et al., 2006; Okuyemi et al., 2004; Gandhi et al., 2009).  However, data regarding age of onset of regular smoking are mixed. Six studies found no difference (Hyland et al., 2002; Okuyemi et al., 2004; Okuyemi et al., 2007; Cubbin et al., 2010; Stahre et al., 2010; Lawrence et al., 2010 – males only), two found that menthol smokers began regular smoking at a later age (Fagan et al., 2010; Fernander et al., 2010), and two found that menthol smokers began regular smoking at an earlier age (Lawrence et al., 2010 – females only; Nonnemaker secondary data analysis).”

Nicotine Dependence

The report concluded that “…menthol in cigarettes is likely associated with increased dependence.”  This position defies some key research – “[Cigarettes per day] and [the Fagerström Test for Nicotine Dependence], two measures that have historically been used to assess nicotine dependence, find no consistent effect of menthol” – and appears to be largely “based on findings of TTFC [time to first cigarette]” and other scales of dependence.  Let’s look at the “findings of TTFC.”

One of the main sources for such data is a 2010 study by Fagan) which found that only moderate menthol smokers (6-10 cigarettes per day, or cpd) were 22 percent more likely than regular smokers to have a cigarette within 5 minutes of waking.  Strangely, menthol had no effect among light (< 5 cpd) and heavy (11+ cpd) smokers.  A study by Collins and Moolchan in 2006 reported a higher percentage of 5-minute TTFC among adolescent menthol smokers seeking cessation treatment, a highly selective population that is not representative of all menthol smokers.  The FDA dismisses the results from four other studies, some positive and some negative, because they didn’t analyze the first five minutes after waking, which the FDA apparently views as the standard.  The report also discusses several studies that are not peer-reviewed, even though “…more consideration was given to peer-reviewed studies.”

The FDA conclusion conflicts with a 2012 study that was not included in the report (available here). It found that “Smoking behaviors may vary by menthol, but menthol was unassociated with dependence.”

Smoking Cessation

The FDA report concluded that “…menthol in cigarettes is likely associated with reduced success in smoking cessation, especially among African American menthol smokers.”  Once again, the finding was based on decidedly mixed results.

The report discussed nine cohort studies that had a vast range of outcomes.  Three found no differences in cessation between menthol and nonmenthol smokers (Hyland et al., 2002; Cropsey et al. 2009; Murray et al., 2007), and the report criticized two of them.  “A fourth study (Blot et al., 2011) found no difference between African American smokers but that White menthol smokers were more likely to have quit.  Of the remaining five cohort studies, four found worse cessation outcomes for menthol smokers compared to their nonmenthol counterparts ([only three references were cited here] Pletcher et al.,2006; Okuyemi et al., 2003; Harris et al., 2004), and one had a trend towards menthol smokers having worse outcomes (Foulds et al., 2006).”

Six cross-sectional studies had varied results regarding menthol and cessation: “…two (Fu et al., 2008; Muscat et al., 2002) failed to find significant differences between menthol and non-menthol smokers [The report appears to cite the wrong study by Fu et al., this one by the same author is more relevant  ]…three [studies] found that menthol smokers had worse cessation outcomes as compared to their nonmenthol smoking counterparts [references were not provided], while one (Gundersen et al., 2009), found that African-American and Latino menthol smokers had worse cessation outcomes as compared to their nonmenthol smoking counterparts while the reverse was true for White smokers.” 

The above FDA description of the Gundersen study is a disturbing misappropriation of research results.  The agency implies that the cessation outcome was worse in both African-American AND Hispanic menthol smokers.  Hispanic menthol smokers had worse cessation outcomes (adjusted odds ratio, AOR= 0.61, 95% Confidence Interval, CI=0.39-0.97), but the cessation rate for African American menthol smokers was NOT significantly lower (AOR=0.78, CI=0.56-1.09).  Statistical significance was reached only when both groups were combined as “Non-White”.  In fact, Gundersen describes the overall result for all menthol smokers: “The odds of being a former smoker does not differ statistically or substantially relative to nonmenthol smokers.”

The bottom line: Numerous studies showed mixed results for menthol and cessation, with some suggesting that menthol might be associated with reduced cessation among African Americans and with increased cessation among whites. 

The FDA preliminary evaluation – reflecting data from numerous studies – does not provide evidence of any significant differences between menthol and regular cigarettes with respect to smoking initiation, addiction to nicotine or cessation.  There is no justification for an evidence-based decision by the FDA to ban or otherwise restrict the menthol content in cigarettes.