Why Smoke Flavor Capsule Cigarettes? Preferences and Perceptions Among Adult Smokers in Mexico

Abstract

Introduction

Flavor capsule cigarettes are rapidly gaining global market share, particularly in Latin America; however, little is known about consumer profiles and perceptions since the diffusion of this product innovation has become more widespread.

Aims and Methods

This paper compares profiles and perceptions of smokers who prefer flavor capsule cigarettes and those who do not in Mexico. We analyzed data from six online surveys of adult Mexican smokers (n = 4488) from 2018–2020. Participants identified their preferred brand variety characteristics (no capsule; one capsule; two capsules), and reported perceptions of their brand varieties’ flavor, smoothness, harmfulness, and whether they smoke to control their appetite. Separate multinomial logistic models regressed preferred cigarette type on sociodemographic and smoking-related variables, as well as perceived benefits on preferred cigarette type, adjusting for covariates.

Results

Most participants preferred capsule cigarettes (single capsule = 47.5%; double capsule = 12.9%). Flavor capsule users reported that their preferred varieties mostly tasted like menthol/mint (59% of single capsule users and 23% of double capsule users), cucumber (12% and 27%, respectively) or berries (10% and 22%, respectively). Females were more likely than males to prefer either single or double capsule cigarettes (AOR = 2.26 and 2.01, respectively). Preference for flavor capsules was also higher among younger smokers, e-cigarette users, those who smoke less frequently, and those who recently tried or plan to quit. Smokers who preferred capsules were more likely than non-capsule smokers to perceive their cigarettes as smoother, less harmful, and to smoke to control their appetite.

Conclusions

The high prevalence of capsule use and widespread perceptions of the benefits of capsules over traditional cigarettes may help explain why tobacco control policies have not reduced smoking prevalence in Mexico.

Implications

Flavor capsules are rapidly growing in popularity in many countries around the world, including Mexico. This study found that more Mexican adult smokers prefer brand varieties with capsules (60%) than without capsules (40%), that flavor capsule use strongly correlates with menthol cigarette use, and that perceived flavors of capsules range widely. Flavor capsule smokers have more positive smoking expectancies (eg, lower risk perceptions, dietary control) than noncapsule smokers. Capsules appeal particularly to women, young adults, less frequent smokers, and those who recently tried to quit. Policies should consider banning flavors in cigarettes, including capsules and other filter innovations.

Introduction

Flavor capsule cigarettes are a relatively recent tobacco product innovation that has rapidly captured market share around the world, particularly in Latin America.^1–3 As of 2017, the global capsule market was estimated to be 150 billion sticks.¹ These cigarettes contain a capsule in the filter that users crush to release a liquid that flavors the smoke.⁴ The variety of flavors in the capsules is rapidly expanding, moving beyond traditional menthol to include diverse fruit flavors and cocktails.¹ As the market has grown, industry has also introduced cigarettes with multiple capsules and flavors in a single stick, as well as packs with sticks that have different flavors, including flavors randomly allocated across the sticks in the pack, so consumers can guess the flavor. Research on early adopters of flavor capsules indicates their appeal for women, young adults, and lighter smokers,^2,5–9 but little is known about consumer profiles and perceptions since the diffusion of this product innovation has become more widespread. This is an important issue to address because the rapid expansion of the flavor capsule market segment threatens to undermine the World Health Organization’s Framework Convention on Tobacco Control (WHO-FCTC), as appears to have happened in Mexico.^5,7,10 This study compared Mexican adult smokers who prefer flavor capsule cigarettes with those who use regular cigarettes, including a comparison of their expectancies from smoking (eg, relative risk perceptions, appetite suppression). We also evaluated the range of flavors that consumers perceive in the capsules from their preferred brand varieties.

The Latin American region has made important advances in tobacco control, with many countries adopting key policies that the WHO-FCTC recommends.¹¹ Overall, the volume of cigarette sales in the Latin America region has declined, yet industry reports show that sales of flavor capsule cigarettes have increased across the region.³ Indeed, the top 5 countries with the largest market share for flavor capsules are in Latin America and include Mexico.^1,2 Mexico has implemented many of the WHO-FCTC-recommended policies, such as taxes,¹² relatively large pictorial warning labels,¹³ marketing bans (except at point of sale and in adult venues and print media for adults),¹⁴ and partial smoke-free policies.¹⁵ However, adult smoking prevalence did not decline from 2009 to 2016, which is the period over which these policies were implemented (16.5% to 19% overall; 25.8% to 29.4% among males; 8.2% to 9.3% among females), although smoking frequency has decreased.⁷ In 2018, over half (56.9%) of Mexican smokers smoked less than daily, and daily smokers averaged 6 cigarettes a day.¹⁶ Flavor capsules may particularly appeal to these relatively light smokers,¹⁷ keeping them from quitting and helping to explain why the prevalence of smoking has not decreased,¹⁰ as the rapid growth of the flavor capsule market is the most visible industry response to Mexico’s implementation of FCTC policies.^18,19

Adolescents and younger smokers generally find flavor capsules more appealing than older smokers. In Australia, the United States (US), United Kingdom (UK), and Chile, flavor capsule consumption is most common amongst younger adult smokers,^2,5,9,16 with a similar but weaker pattern in Mexico.⁵ While more research is needed on this topic, flavor capsules may be relatively more appealing to older Mexican smokers than older smokers in other countries because of the persistent pattern of light smoking across all age groups, as well as the discount cost for the most popular flavor capsule brand family (Pall Mall) in the country.⁵ Indeed, this combination of low price, international prestige, and value added from cigarette capsules likely explains the rapid rise of Pall Mall in Mexico.^5,10 Finally, studies in the US,⁵ Mexico,⁵ and New Zealand⁶ found that females were more likely than males to prefer flavor capsules, while other studies in the US,¹⁷ Australia,⁵ and the UK¹⁴ found no significant differences by sex.

Both qualitative and quantitative studies provide further insights into the appeal of flavor capsules. For example, focus groups with young females in Scotland and surveys among smokers in Australia, Mexico, and the US, have indicated that smokers perceive flavor capsule cigarettes to be smoother and less harmful than standard cigarettes.^5,20 Another appeal of flavor capsules is their novelty, as found in qualitative studies with young nonsmokers, occasional smokers, and older smokers who viewed flavor capsules as a “cool invention”, a “designer cigarette” and an attempt to capitalize on the popularity of e-cigarette flavors, “put(ting) that into a cigarette”.^8,20 Users report that “bursting” the capsule, including the clicking sound this makes, is appealing when compared to menthol tobacco cigarettes.²⁰ Experiments that systematically manipulate product and packaging characteristics find that flavor capsules appeal to young adult nonsmokers in New Zealand who are susceptible to smoke,⁶ as well as to middle school youth in Mexico.^4,21

As countries grapple with the rapid growth of the flavor capsule market segment, research is needed to inform their regulation, including efforts to ban them. The current study provides recent data on adult smokers in Mexico, where prior research evaluated smokers’ use and perceptions of capsules during their initial introduction into the market (2012–2014).⁵ Since then, the market for flavor capsules has grown,^1–3 with a 2018 national survey indicating that 43% of Mexican smokers used capsules in the last month.⁷ We examine the range of flavors in the brand varieties that capsule smokers prefer, as well as how smokers’ perceptions of smoking benefits vary as a function of their preferences for flavor capsules and menthol tobacco cigarettes.

Methods

Design and Study Population

The data for this study of Mexican adults (18 years old and higher) who were current smokers (smoked in the last 30 days) are from the Mexico administration of International Tobacco Control (ITC) Smoking and Vaping Survey.²² In each of five online surveys, separated by four months (November 24–December 10, 2018; March 16–April 8, 2019; July 17–August 15, 2019; November 20–December 5, 2019; March 16–26, 2020), 1500 participants were recruited through an online consumer panel for marketing research (ie, Kantar). For each survey, quotas were used for educational attainment (eg, at least 500 with high school education or less) and for an oversample of current e-cigarette users (at least 500 who used e-cigarettes in the last 30 days), to ensure adequate sample size for evaluating these groups. Participants were recruited by email from the provider’s existing panel, screened for eligibility, and compensated upon survey completion using a point system. Our open cohort design includes following up participants from one survey to the next, with samples replenished to survey approximately 1500 participants at each wave; however, the analytic sample for the present study includes data only from the first survey to which participants responded (wave 1 n = 1501; wave 2 n = 1035; wave 3 n = 799; wave 4 n = 703; and wave 5 n = 631; Total n = 4669). Participants with any missing data (n = 181) were excluded from the analytic sample, yielding a total analytic sample of 4488. Respondents provided consent prior to starting the survey and took approximately 22 minutes (median) to complete it. All study procedures were approved by the Institutional Review Board and Ethics Committee of the National Institute of Public Health of Mexico (CI 1572).

Measures

Preferred Brand Variety and Perceived Capsule Flavors:

Participants first indicated whether they had a preferred brand variety, with subsequent questions on brand preferences referencing either their “preferred brand” or, in the case of not having one, “the brand they last purchased.” They were asked to identify their preferred brand/last purchase from a list of 25 brands available in Mexico at the time of data collection. Then, they were shown and asked to select the preferred/last purchased brand variety from photos of known pack varieties within the brand family they selected. Then, participants were asked if their preferred/last purchased brand variety included no flavor capsules, one flavor capsule, or two or more flavor capsules. Those who indicated that their preferred or last purchased brand variety (hereafter termed “preferred”) included flavor capsules were next asked to report the flavor they perceived when they crushed the flavor capsule, choosing as many flavors as they liked from a list (menthol or mint; clove; vanilla; berries, for example, strawberry, blackberry, cherry; lemon or citrus flavor; tropical fruits, for example, mango, pineapple; cucumber; alcoholic drinks, such as piña colada, margarita or other cocktails; other flavors). This was done because pack and advertising descriptors for flavor capsules generally do not use flavor names but “concept descriptors” (eg, Havana Nights, Fusion Blast).²³ Also, we coded the pack selected as their preferred brand variety as either regular or menthol tobacco cigarettes.

Brand Perceptions:

Participants reported whether their preferred brand variety was more, less, or equal in smoothness and, separately, harmfulness “compared to other cigarettes.” Responses were categorized in 3 level variables for smoothness (“much harsher” or “somewhat harsher”=reference; “the same” or “don’t know”; “somewhat smoother” or “much smoother”), harmfulness (“more harmful or much more harmful”=reference; “no difference” or “don´t know”; “much healthier” or “healthier”). Finally, participants were asked if they smoke to control their appetite, with responses classified into 3 categories (“it doesn´t apply to me at all” =reference), (“it moderately applies to me/it applies a little to me”) and (“it perfectly applies to me/it applies a lot to me”). This question was assessed to explore whether the sweet flavors in capsules better substitute for eating than regular cigarettes.

Tobacco and Other Substance Use Behaviors and Intentions:

Participants were asked about smoking frequency in the last 30 days and classified into: a) nondaily smoker; b) daily smoker, <= 5 cigarettes per day; and c) daily smoker > 5 cigarettes per day. This median cut point for daily smoking frequency is similar to other samples of Mexican smokers, including nationally representative samples,^24,25 with this categorization indicating predictive validity for predicting cessation behaviors.^25,26 E-cigarette use frequency in the last 30 days was queried to classify participants as follows: a) exclusive smoker; b) occasional dual user (uses e-cigarettes less than three days a week); c) frequent dual user (uses e-cigarettes three days per week or more often). Other smoking-related variables queried included having attempted to quit smoking in the prior four months (yes vs. no) and intentions to quit smoking within the next six months (yes vs. no). Smoking dependence was measured using 10 questions from 5 dimensions of the Brief Wisconsin Inventory of Smoking Dependence Motives (WISDM) scale (eg, I often smoke without thinking about it; alpha = 0.92), which assesses motivational influences on smoking.²⁷ Items were selected based on pilot research indicating they could discriminate between relatively low levels of smoking frequency among smokers of Mexican ancestry in the US, as Mexicans and Mexican Americans have relatively light smoking patterns, with less than half smoking daily.⁷

Sociodemographics:

Participants reported their sex (male or female), age (ie, 18–29, 30–39, 40–49 or 50+ years), highest educational attainment (middle school or less; high school, technical studies or some college; college degree or postgraduate studies), and monthly household income in Mexican pesos (<$8000; from $8001 to $15 000; from $15 001 to $20 000; >$20 000; missing), where the exchange rate was approximately $20 pesos to $1 US dollar.

Statistical Analyses

Using STATA version 15,²⁸ descriptive frequencies were assessed for sample characteristics and, among flavor capsule users, the flavors they reported perceiving when crushing the capsule, reported separately for single and double capsule users. Multinomial models regressed preferred cigarette type (no capsules=reference group; one capsule; two capsules) on sociodemographics, and tobacco-related variables. Multinomial models also regressed trichotomized brand perceptions (ie, perceived harm, perceived smoothness, smoking to control appetite) on preferred cigarette type, adjusting for the same covariates from the multinomial models. In fully adjusted models for all three outcomes, we assessed interactions between preference for menthol cigarettes and for capsules using global F-tests for the set of interaction terms. In the models for the appetite outcome, we also assessed interactions between flavor capsule use and sex to assess whether this perception may explain why women have been found to be more likely to prefer capsule varieties. Sensitivity analyses were conducted using post-stratification weights based on national benchmarks for the sex, age, education, and vaping status profile for the general population of Mexican smokers.^29,30 Results from these analyses were consistent with the unweighted analyses we report here and would not have significantly altered our primary conclusions; hence, we present the unweighted results. Similarly, we conducted sensitivity analyses after excluding individuals who indicated “don’t know” for the brand perception outcomes, and the results (See Appendix Table A) and their interpretation are consistent with what we present here.

Results

Sample Characteristics

Table 1 shows the characteristics of the analytic sample (n = 4488). The mean age of participants was 35.4 years, with those aged 18–29 accounting for more than one-third (37.3%) of the sample. About half (51.7 %) were male and one-third (33.1%) had a university degree or higher. More than half smoked less than daily (55.4%), 43.9% had attempted to quit in the prior four months, and 43.4% concurrently used e-cigarettes (15.7% frequent dual-use; 27.7% occasional dual-use).

Table 1.

Characteristics of Mexican Adult Smokers, 2018–2020

Characteristics	%(n = 4488)
Sex
Female	48%
Male	52%
Age group
18-29	37%
30-39	31%
40-49	16%
50 +	16%
Education
Middle school or less	9%
High school/some college	57%
University or more	33%
Household income
≤ 8000	25%
8000–15 000	28%
15 000–20 000	17%
> 20 000	26%
missing	5%
Smoking frequency
Nondaily	55%
Daily <= 5 cigarettes/day	21%
Daily > 5 cigarettes/day	24%
Preferred cigarette type
No capsule	40%
Single capsule	48%
Double capsules	13%
Have a preferred brand
Yes	85%
Cigarette flavor
Regular	60%
Menthol	40%
Smoking dependence
Mean (S.E.)	2.7 (0.94)
Quit attempt (last 4 months)
Yes	44%
Plans to quit in the next 6 months
Yes	38%
Brand smoother than other brands
Somewhat harsher/much harsher	15%
Same	22%
Healthier/Much healthier	63%
Brand less harmful than other brands
More harmful/much more harmful	5%
Same	82%
Healthier/Much healthier	13%
Smokes to control appetite
It doesn´t apply to me at all	52%
It applies a little/moderately to me	33%
It perfectly applies/it applies a lot to me	15%
E-cigarette use
Exclusive smoker	57%
Occasional dual user (<3 days/week)	28%
Frequent dual user (3+days/week)	16%

Overall, 40% of smokers reported preference for menthol flavored tobacco cigarettes, which was less prevalent among smokers who preferred cigarettes without flavor capsules (9.8%) than those who prefer either single or double flavor capsule varieties (59.3% and 58.7%, respectively; data not shown).

Prevalence and Correlates of Capsule Use

Most of the sample (60.4%) indicated that their preferred brand variety included a capsule, with 47.5% preferring single capsule and 12.9% preferring double capsules. Fully adjusted multinomial regression models (Table 2) indicated that females were more likely than males to prefer varieties with single capsule (Adjusted Odds Ratio [AOR] = 1.79; 95%CI = 1.53, 2.09) or double capsule (AOR = 1.61; 95%CI = 1.30, 2.00). Preference for either single or double flavor capsule cigarettes was inversely associated with age, with results indicating a lesser likelihood across age categories (Table 2). Smokers who preferred menthol tobacco cigarettes were more likely to prefer cigarettes with single (AOR = 14.04; 95%CI = 11.61, 16.98) and double capsules (AOR = 13.22; 95%CI = 10.36, 16.86) than cigarettes without flavor capsules. Compared to nondaily smokers, preference for flavor capsules was lower among daily smokers who smoke more than 5 CPD (AOR = 0.75 for single capsules; AOR = 0.54 for double capsules). Smokers who attempted to quit in the prior four months were more likely to smoke double capsules than smokers who did not attempt to quit (AOR = 1.41; 95%CI = 1.11, 1.79 for double capsules). Compared to exclusive smokers, occasional dual users (AOR = 1.43; 95%CI = 1.19, 1.72) or frequent dual users (AOR = 2.66; 95%CI = 2.10, 3.35) were more likely to prefer varieties with a single capsule, but only dual users who used e-cigarettes frequently were more likely to prefer double capsules (AOR = 1.86, 95%CI = 1.33, 2.59).

Table 2.

Multinomial Regression of Preference for Flavor Capsule Cigarettes vs Cigarettes Without Capsules on Sociodemographics & Smoking-Related Variables

Variables	No capsule	Single capsule					Double Capsule
	%	%	OR	95% CI	AOR1	95% CI	%	OR	95% CI	AOR1	95% CI
Sex
Male	47.9%	40.7%	Reference				11.5%	Reference
Female	30.6%	54.9%	2.10	1.85, 2.39**	1.79	1.53, 2.09**	14.5%	1.97	1.63, 2.38**	1.61	1.30, 2.00**
Age
18–29	32.7%	50.4%	Reference				16.9%	Reference
30–39	35.9%	50.7%	0.91	0.78, 1.07	0.95	0.78, 1.15	13.4%	0.72	0.57, 0.90*	0.83	0.64, 1.07
40–49	47.1%	41.9%	0.57	0.47, 0.69**	0.57	0.45, 0.72**	11.1%	0.45	0.34, 0.60**	0.46	0.33, 0.64**
50 +	54.5%	40.8%	0.48	0.40, 0.58**	0.68	0.53, 0.86*	4.8%	0.17	0.11, 0.24*	0.23	0.15, 0.35**
Education
Middle school or less	46.1%	37.1%	Reference				16.8%	Reference
High school/some college	40.2%	46.0%	1.42	1.13, 1.78*	1.21	0.92, 1.61	13.9%	0.94	0.70, 1.27	0.85	0.59, 1.21
University or more	36.7%	53.1%	1.80	1.42, 2.28**	1.79	1.31, 2.45**	10.2%	0.76	0.55, 1.06	0.95	0.63, 1.44
Household income
≤ 8000	36.5%	46.6%	Reference				16.9%	Reference
8000 - 15000	40.0%	47.6%	0.93	0.78, 1.11	0.8	0.64, 0.99*	12.5%	0.67	0.52, 0.86*	0.68	0.51, 0.90*
15000-20000	40.2%	47.9%	0.93	0.76, 1.13	0.73	0.56, 0.94*	11.9%	0.63	0.47, 0.85*	0.67	0.48, 0.95*
> 20000	41.7%	48.7%	0.91	0.76, 1.09	0.71	0.56, 0.91*	9.7%	0.49	0.38, 0.65**	0.59	0.42, 0.83*
missing	38.8%	44.5%	0.89	0.64, 1.24	0.86	0.58, 1.29	16.8%	0.93	0.60, 1.43	0.91	0.54, 1.52
Smoking frequency
Nondaily	35.1%	49.2%	Reference				15.7%	Reference
Daily <= 5 cigs	38.5%	49.6%	0.91	0.78, 1.08	0.95	0.79, 1.19	11.9%	0.68	0.54, 0.87*	0.86	0.66, 1.17
Daily > 5 cigs	50.9%	41.7%	0.58	0.50, 0.68**	0.75	0.61, 0.93*	7.4%	0.32	0.24, 0.42**	0.64	0.46, 0.89*
Have a preferred brand
No	36.6%	47.8%	Reference				15.7%	Reference
Yes	40.1%	47.5%	0.90	0.75, 1.08	0.8	0.64, 1.00	12.4%	0.72	0.56, 0.93*	0.78	0.58, 1.04
Cigarette flavor
Regular	59.0%	32.4%	Reference				8.6%	Reference
Menthol	9.5%	72.0%	13.35	11.1, 16.0**	14.04	11.6, 17.0**	18.5%	13.02	10.3, 16.4**	13.22	10.4, 16.9**
Smoking dependence	2.8 (0.92) a	2.8 (0.96)	1.03	0.97, 1.11	1.07	0.98, 1.18	2.5 (0.9)	0.73	0.66, 0.82**	0.87	0.76, 0.99*
Quit attempt (last 4 months)
No	42.9%	45.2%	Reference				11.9%	Reference
Yes	35.2%	50.6%	1.36	1.20, 1.55**	1.14	0.96, 1.36	14.2%	1.46	1.20, 1.76**	1.41	1.11, 1.79*
Plan to quit (next 6 months
No	41.7%	45.0%	Reference				13.2%	Reference
Yes	35.9%	51.6%	1.33	1.16, 1.51**	1.18	0.99, 1.41	12.4%	1.08	0.89, 1.32	0.95	0.74, 1.21
E-cigarette use
Exclusive smoker	44.7%	49.2%	Reference				13.7%	Reference
Occasional dual user(<3 days/week)	36.9%	49.6%	1.49	1.28, 1.72**	1.43	1.19, 1.72**	11.9%	1.05	0.84, 1.31	1.01	0.78, 1.30
Frequent dual user (3+days/week)	27.2%	41.7%	2.41	1.99, 2.92**	2.66	2.10, 3.35**	11.5%	1.38	1.03, 1.84*	1.86	1.33, 2.59**

Multinomial regression models, adjusted for variables specified in the models (147 missing data in family income variable).

Mean and SD

* p-value < .05, **p-value <.01

Table 3 shows associations between cigarette preferences and perceived smoothness, perceived lower harm, and smoking to control appetite. Compared to those who prefer cigarettes without capsules, those who preferred single or double capsule varieties were more likely to report that their brand variety was either equally smooth (AOR = 2.51, 95% CI = 1.98, 3.18; AOR = 4.64, 95%CI = 2.83, 7.59, respectively) or smoother than other cigarettes (AOR = 6.70, 95% CI = 5.42, 8.27; AOR = 13.42; 95%CI = 8.49, 21.22, respectively). Preference for menthol tobacco cigarettes was also positively associated with perceptions of greater smoothness compared to other cigarettes (AOR = 5.07; 95%CI = 4.04, 6.38).

Table 3.

Preferred Cigarette Type, Menthol Use, Brand Perceptions, and Functional Use of Smoking

Dependent variable & preferred cigarette type	%	%	OR	(95% CI)	AOR1	(95% CI)	%	OR	(95% CI)	AOR1	(95% CI)
Smoothness compared to other cigarettes	Somewhat or much harsher (Ref)	Same					Somewhat or much smoother
Capsule preference
Without capsule	29	28	Reference				43	Reference
Single Capsule	7	18	2.57	(2.05, 3.22)**	2.51	(1.98, 3.18)**	75	7.01	(5.73, 8.58)**	6.7	(5.42, 8.27)**
Double Capsule	4	18	5.02	(3.09, 8.15)**	4.64	(2.83, 7.59)**	78	14.23	(9.06, 22.36)**	13.42	(8.49, 21.22)**
Cigarette flavor preference
Regular	54	25	Reference				22	Reference
Menthol	77	17	2.54	(1.97, 3.28)**	2.41	(1.86, 3.11)**	6	5.26	(4.20, 6.59)**	5.07	(4.04, 6.38)**

Harmfulness compared to other cigarettes	Somewhat or much more harmful (Ref)	Same					Healthier or much healthier
Capsule preference
Without capsule	5	87	Reference				7	Reference
Single Capsule	5	78	1.04	(0.77, 1.39)	1.18	(0.86, 1.60)	17	2.73	(1.92, 3.86)**	2.77	(1.92, 4.00)**
Double Capsule	5	84	0.97	(0.63, 1.48)	1.08	(0.69, 1.69)	11	1.47	(0.88, 2.46)	1.79	(1.05, 3.06)*
Cigarette flavor preference
Regular	13	82	Reference				5	Reference
Menthol	13	83	1.15	(0.86, 1.54)	1.14	(0.85, 1.53)	5	1.12	(0.81, 1.56)	1.14	(0.85, 1.53)

Smoke to control appetite	Doesn´t apply to me at all (Ref)	Applies a little or moderately to me					Applies perfectly or a lot to me
Capsule preference
Without capsule	58	30	Reference				12	Reference
Single Capsule	47	36	1.46	(1.27, 1.68)**	1.37	(1.17, 1.61)**	18	1.85	(1.53, 2.24)**	1.77	(1.38, 2.27)**
Double Capsule	58	30	0.99	(0.80, 1.23)	1.14	(0.90, 1.44)	12	0.99	(0.73, 1.34)	1.44	(0.98, 2.10)
Cigarette flavor preference
Regular	51	35	Reference				15	Reference
Menthol	54	31	0.86	(0.75, 0.99)*	0.91	(0.78, 1.06)	15	0.98	(0.82, 1.17)	1.1	(0.88, 1.39)

Models adjust for sex, age, education, income, frequency of smoking, e-cigarette use, nicotine dependence, quit attempts, quit intentions.

*p < .05; **p < .01.

Compared to noncapsule smokers, those who preferred flavor capsules perceived their brand as healthier than other brands, whether they preferred one capsule (AOR = 2.77, 95%CI = 1.92, 4.00) or two (AOR = 1.79, 95% CI = 1.05, 3.06). Those who preferred single capsule cigarettes were also significantly more likely than noncapsule smokers to report that they smoke to control appetite, whether reporting that this motivation for smoking applies to them either moderately well (AOR = 1.37; 95%CI = 1.17, 1.61) or very well (AOR = 1.77; 95%CI = 1.38, 2.27). In the fully adjusted model for smoking to control appetite, the interaction between sex and cigarette preference was not statistically significant. Furthermore, no significant interactions were found between cigarette flavor (regular vs. menthol) and capsule preference (none, single, double) for the three perception outcomes when conducting a global interaction test for the full models (perceived smoothness F-test = 0.34; perceived lower harm F-test = 0.43; smoking to control appetite F-test = 0.98)

Figure 1 reports flavor capsule smokers’ perceptions of the flavors in the capsule(s) for the variety they prefer. For those who preferred either single or double capsule varieties, the most prevalent flavors were mint/menthol (59% and 23%, respectively), cucumber (12% and 27%, respectively), and berries (10% and 22%, respectively).

Figure 1.

Perceived flavors in capsules among adult smokers who prefer single and double capsule varieties*. *From information of brand variety whether preferred or last purchased from the respondents, they were asked if their preferred/last purchased brand variety included no flavor capsules, had a single capsule, or double flavor capsule. They were also asked to report the flavor they perceived when they crushed the flavor capsule, they could choose as many flavors as they liked from a list.

Discussion

The prevalence of flavor capsule cigarette users in our 2018–2020 sample of Mexican adult smokers (60.4%) is even higher than among smokers in a 2018 national sample of Mexicans 10 and older (43%),³¹ and it is much higher than in a 2014 sample (14%) that used the same consumer panel provider as in our study.⁵ This increase is consistent with industry reports on the rapid growth of flavor capsules in Latin America, including Mexico.^1,2 Our oversample of smokers who concomitantly use e-cigarettes, or “dual users” (43.4% of the sample), which was only 4–5% of adult smokers in a 2016 national survey in Mexico, may have over-represented capsule use, as dual-use was positively associated with preference for capsules, perhaps because of the appeal of a variety of flavors. Our findings suggest that preferences for flavor capsules continue to grow, consistent with arguments that industry expansion of novel flavored cigarette varieties in the Mexican market explain why the population-prevalence of smoking did not decrease after Mexico implemented many FCTC-recommended policies.⁷ The appeal of these products for relatively infrequent smokers—who comprise the majority of smokers in Mexico—may help explain why people keep smoking even as smoking frequency has declined. However, we also found that smokers who recently tried to quit differentially preferred flavor capsules, raising the possibility that their use could lead to cessation. As our cross-sectional study design is limited, longitudinal studies are needed to better evaluate whether and how this product innovation may account for why smoking prevalence remains unchanged despite the continued decrease in the already low frequency of smoking—a key indicator of nicotine dependence. Given that declines in smoking have stalled for those who prefer menthol cigarettes in the US, the strong association we found between Mexicans’ preferences for capsules and menthol tobacco cigarettes is also concerning.³²

Our findings that capsules appeal to women more than men is consistent with prior research in Mexico,^5,31 as well as the USA⁵ and New Zealand.⁶ In addition, a study conducted among female nonsmokers and occasional smokers in Scotland indicated that females found flavor capsules appealing largely due to the technology of flavor capsules.²⁰ This pattern of preference raises concerns that flavor capsules could be a tobacco industry strategy to target women, just as industry development and marketing of menthol brands and slim cigarettes aimed to increase smoking initiation among females according to internal industry documents.^20,33 While we did not find that female capsule users were more likely than male capsule users to smoke to control their appetite, further studies should explore the driving factors for using flavor capsules among females.

We found that younger smokers are more likely to use flavor capsule cigarettes than older smokers, in line with findings in other countries.^2,5,8,9 The exception is research among Mexican smokers from 2012 to 2014,⁵ where preferences for capsules were similar across all age groups except the oldest, among whom this preference was lowest. While we found more evidence of an age gradient, our results and other recent research³¹ indicate that preference for flavor capsules has grown significantly across all age groups, with a significant proportion of all groups preferring capsules (range = 45.6% to 67.3% in our study). Further research should evaluate whether this growing popularity of flavor capsules across age groups is due to relatively low loyalty to particular brand varieties, which may be associated with the generally low frequency of smoking in Mexico. Furthermore, given the value added for cigarette varieties with vs. without flavor capsules, the relatively low price for flavor capsule varieties from the dominant Pall Mall brand family^5,10 may also help explain their appeal among smokers from all age groups.

The finding of higher perceived smoothness for flavor capsules than cigarettes without them is in line with findings from early research on smokers from the UK,⁹ Mexico, Australia, and the US,⁵ as well as when comparing former smokers and young adults who were never smokers but were susceptible to smoke.⁶ This perception may be because most flavor capsule cigarettes contain menthol that can mask harshness of smoking.³⁴ Indeed, we found that participants who prefer menthol tobacco cigarettes reported that their cigarettes were smoother than other brands; however, these smokers were also much more likely to prefer single or double capsules than cigarettes without capsules. After implementation of FCTC policies in Mexico, the tobacco industry increased the availability and range of flavored tobacco products,^18,19 including varieties that include both menthol cigarettes and menthol flavor capsules.³⁵ Our study finds that this combination is popular, with each characteristic having an independent effect on perceptions of smoothness. This is concerning because smoking initiation is associated with use of menthol cigarettes, and preference for menthol cigarettes is associated with higher dependence and relapse rates.³⁴ Perceived smoothness may also be enhanced by the other flavors, such as the relatively popular cucumber flavor, which also has a cooling effect and was a popular flavor during the rapid expansion of JUUL e-cigarette devices in the US.³⁶ Pharmacokinetic research on flavor capsules is needed to better understand its additive potential, as the only pharmacokinetic study to date was limited to a small sample of frequent smokers who prefer menthol tobacco.³⁷

The finding that flavor capsule users perceive their cigarettes to be less harmful than nonflavor capsule cigarettes is also consistent with findings from prior research with Mexican adult smokers when capsules were more novel (2012–2014)⁵ and among Mexican adolescents,²⁰ although this perception was not found for young adults in New Zealand.⁶ This finding suggests that the misperception that some cigarettes are less harmful than others is particularly pronounced among flavor capsules users in Mexico, even though we found that preference for menthol cigarettes is unassociated with this misperception. This misperception may be due to sensorial perceptions of greater smoothness, which may be an indicator of “experiential” risk perception and appears distinct from deliberative/cognitive and affective components of perceived risk.³⁸ This experience may be particularly meaningful for less frequent smokers who are less accustomed to the harshness of smoke, as infrequent smoking is the dominant pattern in Mexico,^7,16 which contrasts with patterns found in high income countries. Whatever the case, imbalances in this misperception may provide government with the evidence it needs to correct this misperception, whether by prohibiting the product attribute or regulations on the marketing and labeling of these products.

Smokers who preferred cigarettes with a single capsule, but not double capsules, were more likely to smoke to control their appetite than smokers who preferred noncapsule cigarettes. This has not been evaluated in prior research, suggesting the need for further research on this topic, perhaps through laboratory studies (eg, pharmacokinetics) or real-time data collection (eg, Ecological Momentary Assessment). These studies could investigate whether preference for certain flavors has a stronger effect on appetite suppression, as well as the timing of smoking to control appetite.

In our sample, the top three capsule flavors that smokers identified in the capsule brand they prefer were mint/menthol, cucumber, and berries, though other more exotic flavors were also described (eg, tropical fruits, alcoholic drinks). More research needs to be done on the role these flavors play in smoking initiation, progression, and cessation. For instance, susceptible nonsmokers in New Zealand were more interested in trying cigarettes with Fruit Burst, Pineapple-Mango, and Hawaiian Mojito flavor capsules than smokers.⁶ Furthermore, nondaily smokers in that study were more likely to prefer menthol and fruit burst flavor than daily smokers. These data suggest that flavors may not only attract new smokers, but also work differently across stages of smoking. That never smokers, especially early adolescents and young adults, find flavor capsules appealing is concerning and raises questions about whether banning them would make them less likely to smoke.²⁰ Tobacco flavors, including flavor capsules, have been banned in Canada, Ethiopia, Senegal, Uganda, the European Union, and Brazil.⁹ These bans are likely to make smoking less appealing to youth, although, to our knowledge, the specific impact of banning capsules has not yet been assessed. Tobacco filter innovations that grow market share, appeal to youth, and perpetuate misperceptions of reduced harm should be targets for regulation. The tobacco industry cannot claim that capsules are long-held trade secrets, and therefore arguments against banning them may be weaker than industry arguments that undermine bans of flavors added to the tobacco leaf or paper.

A limitation of the study is that our sample over-represented smokers from higher socioeconomic status groups. These participants might have over-represented capsule users, since capsule use is somewhat more prevalent among higher SES smokers,³¹ including in our sample. Similarly, our results around perceptions of brand smoothness and harm may have been higher than expected if high SES participants are more likely to smoke premium brands that are also associated with perceptions of greater smoothness and lesser harm.^5,21 In addition, our oversampling of e-cigarette users and relatively younger smokers likely skewed the results toward greater capsule use than in the general population. However, the consistency of our results with prior studies, including the high prevalence of capsule use in national samples in Mexico (6.6% prevalence which is 43% of all smokers),³¹ suggests that these biases are unlikely to explain away our results or their interpretation. Given the range of perceived flavors in capsules, future research should evaluate similarities and differences in flavor preferences across products, especially e-cigarettes, and the role these preferences play in trajectories of use.

Conclusions

Flavor capsules continue to be popular in Mexico, particularly among females and younger smokers. The perception of capsules as smoother and the misperception that they are less harmful suggest a need for government action.

Supplementary Material

A Contributorship Form detailing each author’s specific involvement with this content, as well as any supplementary data, are available online at https://academic.oup.com/ntr.

Funding

Research reported in this study was supported by of the Fogarty International Center of the National Institutes of Health under award number R01 TW010652. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Declaration of Interests

The authors certify that there is no actual or potential conflict of interest in relation to this research.

Data Availability Statement

The data underlying this article will be shared on reasonable request to the corresponding author (James F. Thrasher).