• Open Access

Smokers have varying misperceptions about the harmfulness of menthol cigarettes: national survey data


Correspondence to:
Associate Professor Nick Wilson, Department of Public Health, University of Otago, Wellington, PO Box 7343, Wellington South, New Zealand; e-mail: nick.wilson@otago.ac.nz


Objective: To describe the prevalence of menthol use and perceptions of relative harmfulness among smokers in an ethnically diverse population where tobacco marketing is relatively constrained (New Zealand).

Methods: The New Zealand (NZ) arm of the International Tobacco Control Policy Evaluation Survey (ITC Project) utilises the NZ Health Survey (a national sample). From this sample we surveyed adult smokers, with Wave 2 (n=923) covering beliefs around menthol cigarettes.

Results: Agreement with the statement that “menthol cigarettes are less harmful than regular cigarettes” was higher in smokers who were: older, Māori, Pacific, Asian, financially stressed and had higher levels of individual deprivation. Most of these associations were statistically significant in at least some of the logistic regression models (adjusting for socio-economic and smoking beliefs and behaviour). In the fully-adjusted model this belief was particularly elevated in Pacific smokers (adjusted odds ratio [aOR] = 7.36, 95% CI = 1.92 – 28.27) and also in menthol smokers (aOR = 4.58, 95% CI = 1.94–10.78). Most smokers in this study (56%), and especially menthol smokers (73%), believed that menthols are “smoother on your throat and chest”.

Conclusion: Various groups of smokers in this national sample had misperceptions around the relative harmfulness of menthols, which is consistent with most previous studies.

Implications: This evidence, along with a precautionary approach, supports arguments for enhanced regulation of tobacco marketing and tobacco ingredients such as menthol.

The use of ‘light’ and ‘mild’ descriptors on cigarette packaging has been associated with smoker misperceptions around reduced harm to health.1 As a result of this evidence, many countries have responded to the Framework Convention on Tobacco Control recommendations that these descriptors be banned.2 Such an approach has not yet been taken for the descriptor ‘menthol’ and use of this flavouring ingredient has not yet been restricted by regulations, although the FDA in the United States (US) is reviewing data on this ingredient.3

Research has suggested that mentholated cigarettes (‘menthols’) may pose a greater risk in terms of youth uptake4–6 and that menthol smoking is either “at least” as dangerous as smoking non-mentholated counterparts7 or possibly more hazardous.8,9 Indeed, at a conference about menthols in 2009 there were 10 arguments articulated for the banning of this ingredient.10

The existing literature about smokers’ harm perceptions of menthols is limited to US studies (largely involving African Americans). These have provided evidence that at least some smokers perceive menthols as less hazardous to health.11–14 But one recent US study indicated the opposite pattern was more common in smokers, i.e. menthols were perceived as more risky.15 To explore this issue further, we studied menthol use and harm perceptions among New Zealand (NZ) smokers. This country has an ethnically diverse population and is one where advertising and sponsorship of tobacco products is largely restricted.


The ITC Project

The International Tobacco Control Policy Evaluation Survey (ITC Project) is a multi-country study16 with a NZ component. The NZ survey derives its sample from the national NZ Health Survey (NZHS). Respondents were selected by a complex sample design, which included systematic boosted-sampling of the Māori, Pacific and Asian populations (for details see an online report).17


From the NZHS sample of adult smokers who were 18+ years and willing to participate in further research, a total of 1,376 respondents completed a telephone questionnaire. This gave an overall response rate of 32.6% (when considering the response rate to the NZHS, willingness to further participate, and response to the telephone survey; see an online report17 for details). There was subsequent attrition in participation of 32.9% between Waves, leaving 923 respondents in Wave 2.


Data collection was carried out using a computer-assisted telephone survey (Wave 1 followed the NZHS by 3–4 months; and Wave 2 a year later, i.e. between March 2008 and February 2009). The study protocol was cleared by the Multi-Region Ethics Committee in NZ (MEC/06/07/071) and the Office of Research Ethics, University of Waterloo, Canada (ORE #13547).


Respondents were asked about the brand of tobacco they last purchased and this was used to classify them as menthol users or not. Participants were also asked whether they agreed or disagreed with the statement “menthol cigarettes are less harmful than regular cigarettes” and their perceptions of smoothness and menthols. As this question was only asked in the Wave 2 survey, we focused on results for this Wave only.

Weighting and statistical analyses

Weighting of the results was necessary given the sampling design (e.g. boosted sampling of three ethnic groups in the NZHS) and non-response for the NZHS and ITC Project survey (see an online report).18

We conducted a univariate analysis for the association of the demographic, socio-economic and smoking-related variables with the belief that menthol cigarettes were less harmful than other cigarettes. We then conducted a multivariable logistic regression analysis using a conceptual framework that assumed there would be hierarchical relationships between demographic and socio-economic factors19 that would dominate over smoking-related beliefs, intentions and behaviours. All models included age, gender and ethnicity. Model 2 added the two deprivation measures and two measures of financial stress shown in the table (see an online report for extra details).17 Model 3 added key variables relating to smoking beliefs, intentions and behaviour, i.e. concern about smoking impact on health and quality of life in the future (two-item scale); strength of intention of quitting (four-point scale); and a heaviness of smoking index (alternate version).17 All analyses were conducted in Stata (version 10, StataCorp) and all of the presented results were weighted and adjusted for the complex sample design of the NZHS to make the sample representative of all NZ smokers.


Three quarters of the Wave 2 respondents stated that they had previously smoked menthol cigarettes at some time (75.7%; i.e. 699/923), which included current menthol smokers. We focused on this group as it had some experience with the product and asked its response to the statement that “menthol cigarettes are less harmful than regular cigarettes”. To this statement there was partial agreement (9.5%; 66/698), some uncertainty (9.3% [65/698] saying “don't know” or “neither agree nor disagree”), and majority disagreement (81.2%; 567/698) (with one refusal to answer the question). Agreement that menthols were less harmful was higher in smokers who were older, Māori, Pacific and Asian smokers (Table 1). It rose with each level of increasing individual deprivation and financial stress, and was higher in menthol smokers (Table 1). In contrast to the above results for ethnicity there were only minor differences in actual current menthol use by ethnicity (i.e. Māori 15.5%; Pacific 14.5%; European 12.2%), except for Asian participants (at 1.2%). Beliefs were fairly similar by gender, but current menthol use was much higher for women than men (20.0% vs 5.0%, odds ratio (OR) = 4.74, 95% CI = 2.71 – 8.32).

Table 1.  Beliefs of respondents to menthol cigarette harm by demographic and socio-economic characteristics (all the results weighted to adjust for the complex sample design and non-response).
VariableAgree that menthols are less harmful than regular cigarettes (%)Crude odds ratios (OR) for agreement (95% CI)
  1. Notes:

  2. a. Based on NZHS data with the age data collected a few months prior to the ITC Project survey.

  3. b. For more detail on these measures see an online Methods Report.17

Total (n=698)9.5
Age group (years)a
18–24 (n=45)6.81.00 Referent
25–34 (n=172)9.01.35 (0.36–5.13)
35–44 (n=183)7.11.04 (0.27–4.11)
45–54 (n=160)10.71.64 (0.46–5.77)
55+ (n=138)13.32.11 (0.62–7.20)
Men (n=254)11.91.00 Referent
Women (n=444)7.00.56 (0.29–1.05)
European (includes Other) (n=339)6.31.00 Referent
Māori (n=292)12.12.05 (1.00–4.22)
Pacific (n=38)24.64.83 (1.53–15.31)
Asian (n=29)15.52.72 (0.70–10.64)
Deprivation level (small area) (quintiles)b
1&2 (least deprived) (n=60)7.91.00 Referent
3&4 (n=119)6.40.79 (0.18–3.60)
5&6 (n=121)9.71.26 (0.28–5.60)
7&8 (n=156)8.01.00 (0.25–4.12)
9&10 (most deprived) (n=242)12.41.64 (0.41–6.53)
Individual deprivation (NZiDep) scoresb
0 (least deprived) (n=328)7.11.00 Referent
1 (n=130)8.91.27 (0.51–3.17)
2 (n=81)12.81.68 (0.75–4.94)
3–4 (n=95)11.41.72 (0.47–6.03)
5–8 (most deprived) (n=64)14.02.13 (0.80–5.64)
Financial stressb
Unable to pay any important bills on time –“Yes” (n=45) (referent=“no”)10.21.13 (0.17–7.55)
Not spending on household essentials –“yes” (n=165) (referent=“no”)16.72.75 (1.34–5.65)
Current menthol use (based on brand purchase)
Non-menthol smoker (n=480)8.11.00 Referent
Menthol smoker (n=120)14.81.97 (0.96–4.03)

Most of these associations for belief in lower menthol harmfulness were statistically significant in at least some of the three logistic regression models, some of which adjusted for socio-economic and smoking beliefs and behaviour (Table 2). In particular, Pacific smokers were much more likely to have this belief in the fully adjusted model (model 3), as were menthol smokers.

Table 2.  Logistic regression analysis for smoker beliefs that menthol cigarettes are less harmful than regular cigarettes (all the results weighted to adjust for the complex sample design and non-response, n=698).
VariablesAdjusted Odds Ratio (aOR) (95% CI)a
 Model 1 (demographics)Model 2 (+ socio-economic)Model 3 (+ smoking beliefs, intentions and behaviours)
  1. Notes:

  2. a. The adjusted odds ratios (aORs) represent those agreeing with the statement that menthols are less harmful than regular cigarettes relative to those not agreeing or who did not know.

  3. b. For more detail see an online Methods Report.17

  4. c. For these analyses we conservatively included those without a stated brand preference in the non-menthol group.

Age (35–49 vs <35)1.17 (0.51–2.69)1.21 (0.52–2.79)0.99 (0.37–2.67)
Age (50+ vs <35)2.08 (0.96–4.54)2.42 (1.15–5.09)2.09 (0.73–6.01)
Gender (women vs men)0.61 (0.31–1.18)0.65 (0.33–1.30)0.74 (0.30–1.81)
Māori vs European (includes Other)2.42 (1.16–5.05)2.48 (1.21–5.11)1.29 (0.50–3.30)
Pacific vs European5.52 (1.74–17.49)4.89 (1.52–15.72)7.36 (1.92–28.27)
Asian vs European2.56 (0.60–10.90)2.08 (0.48–9.03)3.47 (0.47–25.65)
Area deprivation quintiles (increasing deprivation)0.98 (0.73–1.32)1.15 (0.83–1.59)
Individual deprivation using NZiDep (any deprivation vs nil)1.58 (0.79–3.16)1.40 (0.63–3.10)
Financial stress: Unable to pay any important bills on time0.65 (0.14–3.03)1.94 (0.39–9.60)
Financial stress: Not spending on household essentials2.33 (1.13–4.79)2.01 (0.80–5.04)
Smoking beliefs, intentions & behavioursb
Concern around smoking impact on health and quality of life in the future (2-item scale)0.68 (0.40–1.18)
Strength of intention of quitting (4-point scale)0.76 (0.46–1.27)
Heaviness of smoking index (alternate version)1.10 (0.92–1.32)
Menthol smoker (current)c4.58 (1.94–10.78)

We also found that more than half (56.0%) of all smokers agreed with the statement that “menthol cigarettes are smoother on your throat and chest than regular cigarettes”. Compared with European participants, Māori were significantly more likely to agree with this statement (OR = 2.56, 95% CI = 1.65 – 3.97). Also, menthol users were more likely to agree that menthols are “smoother on your throat and chest” (73.4% vs 52.9% respectively; OR = 2.45, 95% CI = 1.16 – 5.18).


Main findings

This study found that some groups of smokers (particularly menthol users and Pacific smokers) believed that smoking menthols was less harmful than regular cigarettes. The occurrence of such a belief was generally consistent with most of the former studies in the US (see Introduction). This belief can be regarded as a misperception given available evidence (see Introduction).

Such a misperception may arise from residual tobacco marketing in NZ as occurs with advertisements in imported magazines and possibly the internet,20 but it may also arise from perceptions of the smoothness of menthols given our results showing that such a perception was common. This finding on the perception of smoothness was also broadly consistent with a US study11 and the link with health was consistent with the ITC four-country study that found 42% (Australia) to 54% (UK) of smokers agreed with the statement “the harsher the smoke feels in your throat, the more dangerous the smoke is”.21

Our study also found that women were more likely to smoke menthols than men (as reported elsewhere).7 This finding raises concerns around industry marketing being particularly targeted at women for which there is other NZ evidence.20 Such residual marketing is particularly problematic in terms of inequalities, given the very high smoking rates for Māori girls and young women in this country.22


This study involved a sample that (due to non-participation in the NZHS and subsequent attrition by survey wave) could have become less representative of the national population of smokers. It is, therefore, possible that the weighting process (although relatively sophisticated) may not have fully adjusted for these processes. Nevertheless, in terms of comparisons between groups, the remaining sample still contained a reasonable representation of smokers of all ages, differing ethnicities and deprivation levels (Table 1). Another limitation was that the information around menthols was based on the answers to only a few survey questions.

Research and policy implications

Additional in-depth qualitative and ethnographic work that explored smoker perceptions more fully would be ideal (e.g. especially as to why Pacific smokers might consider menthols to be less harmful). For example, there is evidence that African Americans and Hispanics in the US may associate menthol with medicinal properties.14,23

But despite limits with the current evidence around misperceptions and any additional harm to health from adding menthol to tobacco, a precautionary approach by government is desirable. This should see governments move to ban the marketing of menthols and even its use as an ingredient. Indeed, the FDA powers in the US “leave the door open to future rulemaking limiting or banning the ingredient”.24 Another approach could be for the Framework Convention on Tobacco Control to rule on menthol and other key tobacco product ingredients. Alternatively, regulators could just strive for endgame solutions for all types of commercial tobacco products.25


The ITC Project (NZ) team thank the following groups for their support:

  • • the interviewees who kindly contributed their time to answer the survey questions;
  • • the Health Research Council of New Zealand that provided the core funding for this Project; and
  • • other members of our ITC Project (NZ) Team and international partners (see: http://www.wnmeds.ac.nz/itcproject.html).