Cigarette Smoking and the Role of Menthol in Tobacco Use Inequalities for Sexual Minorities

Abstract

Introduction

Cigarette smoking is disproportionately common among sexual minorities. While menthol market share has increased in the past decade, research on menthol cigarette smoking among sexual minorities is scant. For this study, we examined menthol cigarette smoking in the United States, by sex and sexual identity using a nationally representative sample of adults.

Aims and Methods

We pooled data from the 2015–2019 National Survey on Drug Use and Health. We used chi-square tests and multivariable logistic regression models to examine (1) past 30-day cigarette use, (2) past 30-day menthol use, and (3) menthol preference among adults who have smoked cigarettes in the past 30 days for the following groups: heterosexual/straight males, heterosexual/straight females, gay males, gay/lesbian females, bisexual males, and bisexual females. Data were analyzed in September 2020.

Results

Bivariate analyses revealed that prevalence of cigarette smoking and menthol cigarette smoking were higher among sexual minorities compared with heterosexual respondents, and that preference for menthol was highest among lesbian/gay and bisexual females who smoke. Multivariable models showed that compared with heterosexual females who smoke, odds of menthol preference were lower among heterosexual/straight and bisexual males who smoke, and higher among bisexual females.

Conclusions

Our study found that menthol preference was disproportionately high among sexual minorities who smoke—bisexual and gay/lesbian females who smoke in particular. Research is needed to understand why menthol use is common among sexual minorities who smoke.

Implications

Using data from a nationally representative study, this study makes a unique contribution to literature by identifying differences in menthol smoking between sexual minority subgroups. Specifically, we found that menthol smoking was highest among sexual minority females (ie, lesbian/gay and bisexual females) compared with other sexual minorities. These findings highlight the need for targeted interventions to prevent menthol cigarette initiation among sexual minorities, as well as to promote smoking cessation among sexual minorities who smoke menthol cigarettes, with an emphasis on sexual minority females. Future research should examine the impact of a potential menthol ban on these inequalities.

Introduction

Cigarette smoking is disproportionately high among sexual minorities (ie, individuals who identify as gay, lesbian, or bisexual [LGB]),¹ with recent national data demonstrating that 19.2% of LGB adults currently smoke cigarettes, compared with 13.8% of heterosexual/straight adults.² Furthermore, cigarette smoking among sexual minorities has remained stagnant over the past 30 years,³ which is in stark contrast to the dramatic declines in cigarette smoking among the general population. Research suggests that there is significant within group variation among sexual minorities, with certain LGB subgroups—lesbian/gay and bisexual females in particular—smoking at higher rates^4,5 more heavily,⁶ initiating smoking earlier^1,7 and making fewer quit attempts compared with other sexual minorities, as well as their heterosexual counterparts.⁶

Although menthol has increased in market share over the past decade,⁸ research on menthol cigarette use among sexual minorities is scant. Nascent data suggest that menthol cigarette smoking is common among sexual minorities, sexual minority females in particular.^1,9,10 Indeed, a study using data from the 2012–2013 National Adult Tobacco Survey, which stratified analyses by sex, found that menthol use was significantly greater for sexual minority females who smoke compared with heterosexual females, but that there were no differences in menthol use between sexual minority and heterosexual males who smoke.¹ Of note, these studies^1,9,10 were unable to examine differences between gay/lesbian and bisexual adults who smoke, which could be masking additional within group differences that are evident in studies of tobacco use among sexual minorities.^11,12

Among the general population, menthol cigarettes are associated with increased initiation among youth, greater nicotine dependence and decreased smoking cessation among adults.¹³ Therefore, elevated risk for menthol cigarette smoking among sexual minorities may be impeding efforts to reduce tobacco use among this population. Using pooled data from the 2015–2019 the National Survey on Drug Use and Health (NSDUH), this study examines menthol cigarette smoking, by sexual identity and sex.

Methods

Study Sample

We pooled 5 years of data from the NSDUH public use file (2015–2019; N = 214 505). NSDUH is an annual, nationally representative study of individuals in the United States (ages 12 and older) designed to assess substance use and mental health.¹⁴ This secondary analysis of deidentified, publicly available data received a non-human subjects determination from the Rutgers IRB.

Measures

Sexual Identity

Sexual identity was assessed with the following question: “Which one of the following do you consider yourself to be?” Answer choices included, “heterosexual, that is, straight,” “lesbian or gay,” and “bisexual.” Sexual identity was only assessed among adults (ages 18 and older). Participants were also asked whether they are “male” or “female.” We created a combined sexual identity/sex variable for analyses: heterosexual/straight males, heterosexual/straight females, gay males, gay/lesbian females, bisexual males, and bisexual females.

Past 30-Day Cigarette Smoking

All respondents are asked, “have you ever smoked part or all of a cigarette? (yes/no)” Those who answer “yes” are asked, “during the past 30 days, have you smoked part or all of a cigarette?” Answer choices include, within the past 30 days, more than 30 days but within the past 12 months, more than 12 months but within the past 3 years, and more than 3 years ago. All respondents who reported having smoked a cigarette in the past 30 days were categorized as individuals who smoked in the past 30 days.

Past 30-Day Menthol Smoking

All participants who reported smoking cigarettes in the past 30 days were asked, “Were the cigarettes you smoked during the past 30 days menthol?” (yes/no/don’t know). Those who report “don’t know” are coded as missing in the NSDUH public use file. Past 30-day use of menthol cigarettes was defined as menthol use among the entire sample (ie, those who have not smoked in the past 30 days, those who have smoked menthols in the past 30 days and those who have smoked non-menthols in the past 30 days).

Menthol Preference

Preference for menthol was defined as past 30-day use of menthol cigarettes among those who have smoked cigarettes in the past 30 days (vs. use of non-menthol cigarettes in the past 30 days).

Demographic Characteristics and Other Variables

Other variables included age (18–34 vs. 35 and older), race/ethnicity (non-Hispanic White, non-Hispanic Black, non-Hispanic Other, and Hispanic), annual household income (less than $20 000, $20 000–49 999, $50 000–74 999, $75 000, or more), and survey year (2015, 2016, 2017, 2018, and 2019).

Missing Values

In the public use file, those who selected “don’t know” or refused to answer the question for sexual identity were coded as missing (n = 4113; 1.92% of total sample). There were some missing values for past 30-day menthol preference among those who smoke (n = 596; 1.22%) and past 30-day menthol prevalence (n = 596; 0.28%) due to item non-response. We utilized NSDUH’s imputed variables for past 30-day cigarette use and sex, therefore there are no missing values for these variables. Details on imputation procedures for NSDUH can be found elsewhere.¹⁵ Age, race/ethnicity, and annual household income were included in multivariable models and did not contain any missing values.

Statistical Analysis

Analyses were conducted in September 2020 using SVY procedures in Stata/MP, version 16.1 to account for weighting. Unweighted frequencies and weighted prevalence estimates were used to characterize the sample. Data were analyzed using chi-square tests and multivariable logistic regression models to examine the association between sexual identity/sex and (1) prevalence of past 30-day cigarette smoking, (2) prevalence of past 30-day menthol cigarette smoking, and (3) preference for menthol among those who smoked cigarettes in the past 30 days. Multivariable models controlled for age, race/ethnicity, annual household income, and survey year, with heterosexual/straight females as the reference group. Post hoc, adjusted Wald statistics were used to examine intergroup differences.

Results

Overall, 19.71% (95% confidence interval [CI]: 19.42, 19.99) and 7.63% (95% CI: 7.41, 7.86) of the study population reported smoking in the past 30 days and reported past 30-day menthol cigarette smoking, respectively (Table 1). Among those reporting past 30-day cigarette smoking, 39.05% (95% CI: 38.15, 39.95) reported menthol preference. Past 30-day cigarette smoking (34.16%, 95% CI: 32.71, 35.65), past 30-day menthol smoking (18.27%, 95% CI: 17.12, 19.49), and menthol preference (54.03%, 95% CI: 51.41, 56.62) were highest among bisexual females compared with all other groups. In the multivariable models, all groups reported greater odds of past 30-day cigarette smoking compared with heterosexual/straight females (Table 2). For past 30-day menthol smoking, all groups reported greater odds of past 30-day menthol cigarette smoking compared with heterosexual/straight females, except for bisexual males, for whom there was no difference in odds (adjusted odds ratio [aOR]: 1.09, 95% CI: 0.90, 1.31). Compared with heterosexual/straight females who smoked in the past 30 days, heterosexual/straight males (aOR: 0.60, 95% CI: 0.56, 0.64) and bisexual males (aOR: 0.49, 95% CI: 0.39, 0.63) who smoked in the past 30 days reported lower odds of menthol preference; bisexual females who smoked in the past 30 days reported greater odds of menthol preference (aOR: 1.17, 95% CI: 1.04, 1.32).

Table 1.

Weighted Prevalence of Past 30-Day Cigarette Smoking, Past 30-Day Menthol Smoking, and Menthol Preference Among Those Who Smoked in the Past 30 Days, by Sex and Sexual Identity

	Past 30-day cigarette smoking (N = 210 392)	Past 30-day menthol smokinga (N = 209 820)	Menthol preferenceb (N = 47 454)
	Weighted % (95% CI)	Weighted % (95% CI)	Weighted % (95% CI)
Overall	19.71 (19.42, 19.99)	7.63 (7.41, 7.86)	39.05 (38.15, 39.95)
Heterosexual/straight male (n = 93 506)	21.85 (21.44, 22.27)	7.56 (7.24, 7.89)	34.85 (33.62, 36.11)
Heterosexual/straight female (n = 101 879)	16.54 (16.18, 16.90)	6.98 (6.76, 7.22)	42.62 (41.61, 43.65)
Gay male (n = 2348)	26.02 (23.74, 28.44)	11.33 (9.85, 13.00)	44.20 (38.43, 50.13)
Lesbian/gay female (n = 2292)	29.24 (26.67, 31.95)	14.83 (12.96, 16.91)	50.91 (45.78, 56.02)
Bisexual male (n = 2350)	27.31 (25.01, 29.74)	8.65 (7.30, 10.22)	31.98 (27.20, 37.17)
Bisexual female (n = 8017)	34.16 (32.71, 35.65)	18.27 (17.12, 19.49)	54.03 (51.41, 56.62)
Pearson X2	F(4.40, 659.92) = 222.19, p < .001	F(4.18, 626.97) = 135.09, p < .001	F(3.90, 585.24) = 54.14, p < .001

CI = confidence interval.

^aDefined as menthol use among the entire population, not just those who smoked cigarettes in the past 30 days.

^bDefined as menthol use among those who smoked cigarettes in the past 30 days.

Data source: National Survey on Drug Use and Health, 2015–2019 (pooled). Analyses were restricted to these years since sexual identity wasn’t assessed until 2015.

Those who selected “don’t know” or refused to answer the question about sexual identity were coded as missing.

Table 2.

Overall Sample Characteristics and Multivariable Logistic Regression Models of Sexual Identity and Sex Associated With Past 30-Day Cigarette Smoking, Past 30-Day Menthol Smoking, and Menthol Preference Among Those Who Smoked in the Past 30 Days

	Overall sample	Past 30-day cigarette smoking	Past 30-day menthol smokingg	Menthol preferenceh
	Unweighted n (weighted %)	aOR (95% CI)	aOR (95% CI)	aOR (95% CI)
Sexual identity and sex
Heterosexual/straight female	93 506 (46.50)	Reference	Reference	Reference
Heterosexual/straight male	101 879 (48.46)	1.51 (1.46, 1.56)a,c,d,e,f	1.15 (1.09, 1.21)a,c,d,e	0.60 (0.56, 0.64)a,b,c,d,e
Lesbian/gay female	2348 (1.08)	1.90 (1.66, 2.18)a,b	1.95 (1.66, 2.30)a,b,f	1.10 (0.88, 1.38)b,f
Gay male	2292 (0.85)	1.79 (1.57, 2.04)a,b,e	1.64 (1.39, 1.94)a,b,e,f	0.91 (0.73, 1.13)b,f
Bisexual female	2350 (0.90)	2.12 (1.97, 2.28)a,b,d,f	2.18 (2.00, 2.37)a,b,d,f	1.17 (1.04, 1.32)a,b,f
Bisexual male	8017 (2.22)	1.73 (1.52, 1.97)a,b,e	1.09 (0.90, 1.31)c,d,e	0.49 (0.39, 0.63)a,c,d,e
Age
18–34	113 932 (29.88)	Reference	Reference	Reference
35 and older	100, 573 (70.12)	0.72 (0.70, 0.74)	0.56 (0.54, 0.59)	0.52 (0.49, 0.54)
Race/ethnicity
Non-Hispanic White	128 924 (63.86)	Reference	Reference	Reference
Non-Hispanic Black	27 081 (11.87)	0.77 (0.74, 0.81)	2.71 (2.58, 2.85)	14.84 (13.28, 16.57)
Non-Hispanic Other	21 485 (8.23)	0.68 (0.63, 0.72)	1.06 (0.96, 1.16)	1.86 (1.64, 2.10)
Hispanic	37 015 (16.04)	0.54 (0.52, 0.57)	0.97 (0.90, 1.04)	2.16 (1.98, 2.36)
Annual household income
Less than $20 000	42 934 (16.28)	Reference	Reference	Reference
$20 000–49 999	66 946 (29.46)	0.65 (0.62, 0.67)	0.70 (0.66, 0.74)	1.00 (0.93, 1.07)
$50 000–74 999	33 515 (15.99)	0.46 (0.44, 0.49)	0.53 (0.49, 0.57)	0.98 (0.88, 1.09)
$75 000 or more	71 110 (38.27)	0.30 (0.28, 0.31)	0.35 (0.32, 0.37)	0.93 (0.85, 1.02)
Year
2015	43 561 (19.68)	Reference	Reference	Reference
2016	42 625 (19.82)	0.99 (0.94, 1.05)	0.99 (0.94, 1.05)	1.03 (0.95, 1.11)
2017	42 554 (20.03)	0.92 (0.88, 0.97)	0.96 (0.90, 1.02)	1.06 (0.97, 1.16)
2018	43 026 (20.17)	0.89 (0.84, 0.95)	0.95 (0.88, 1.02)	1.12 (1.03, 1.23)
2019	42 739 (20.29)	0.87 (0.83, 0.92)	0.95 (0.89, 1.01)	1.15 (1.05, 1.26)

Post hoc, adjusted Wald statistics were used to examine intergroup differences in menthol cigarette smoking. Models without superscripts indicate non-statistically significant comparisons. Those who selected “don’t know” or refused to answer the question about sexual identity were coded as missing. aOR = adjusted odds ratio, CI = confidence interval.

^aSignificant (p < .05) compared with heterosexual/straight females (reference).

^bSignificant (p < .05) compared with heterosexual/straight males.

^cSignificant (p < .05) compared with gay/lesbian females.

^dSignificant (p < .05) compared with gay males.

^eSignificant (p < .05) compared with bisexual females.

^fSignificant (p < .05) compared with bisexual males.

^gDefined as menthol use among the entire population, not just those who smoked cigarettes in the past 30 days.

^hDefined as menthol use among those who smoked cigarettes in the past 30 days.

Data source: National Survey on Drug Use and Health, 2015–2019 (pooled). Analyses were restricted to these years since sexual identity wasn’t assessed until 2015.

Missingness: sexual identity (n = 4113, 1.92%); past 30-day menthol preference among smokers (n = 596; 1.22%); past 30-day menthol prevalence (n = 596; 0.28%). The remaining variables did not have missing values.

In all multivariable models, adjusted Wald tests showed differences among sexual minority subgroups (Table 2). For past 30-day cigarette smoking, odds of use did not differ between lesbian/gay and bisexual females or between gay and bisexual males. However, odds of use differed between bisexual females and gay and bisexual males. For past 30-day menthol smoking, odds of use did not differ between lesbian/gay and bisexual females but were greater for gay men compared with bisexual men. Similar to past 30-day cigarette smoking, odds of use also differed between bisexual females and gay and bisexual males. For menthol preference, odds of use did not differ between lesbian/gay and bisexual females who smoked in the past 30 days, while odds of use did differ between bisexual females and gay and bisexual males who smoked in the past 30 days. Bisexual males who smoked in the past 30 days reported lower odds of menthol preference compared with bisexual females and lesbian/gay females who smoked in the past 30 days.

Discussion

To our knowledge, this is the first study to report on elevated menthol smoking among bisexual females. Using a nationally representative sample, we found that menthol cigarette smoking was common among sexual minorities who smoke and identified within group differences in menthol cigarette smoking, with bisexual and lesbian/gay females who smoke reporting the highest prevalence of use. These findings align with prior research showing elevated menthol cigarette smoking among sexual minorities^1,9 and studies that have found that sexual minority females, bisexual females in particular, are at a greater risk for cigarette smoking.^5,11 In terms of subgroup differences by sex, our study builds upon research from Johnson et al., which found differences in menthol preference between heterosexual/straight and sexual minority females who smoke, but not heterosexual/straight and sexual minority males who smoke.¹ Of note, Johnson et al., did not separate bisexual from lesbian/gay respondents, which may have masked differences in menthol preference among males. In our study, we were able to distinguish between these two groups and found differences in menthol preference between straight/heterosexual males who smoke and gay males who smoke, as well as differences between straight/heterosexual females and bisexual females who smoke.

Additionally, we found significant associations between sexual identity and sex and menthol smoking, controlling for various demographic characteristics, including race/ethnicity. However, above and beyond controlling for race/ethnicity, future research should examine menthol smoking behaviors among sexual minorities by race/ethnicity, given the extremely high preference for menthol cigarettes among African Americans who smoke (85%),¹⁰ as well as the long history of the tobacco industry targeting the African American community with menthol advertising.¹⁶ Furthermore, research has found great variation in cigarette smoking behaviors by race/ethnicity among sexual minorities, with sexual minority females of color reporting especially high rates of smoking.¹⁷

One possible reason for elevated menthol smoking among sexual minorities could be tobacco industry marketing. Studies that have examined formerly classified tobacco industry documents revealed that the tobacco industry has been interested in the LGB community for decades. Documents show that tobacco companies have engaged with and targeted the LGB community via direct advertising in LGB publications and publications with large LGB readership,^18–20 promoting their products at LGB venues (eg, gay/lesbians bars),^18,20 sponsoring LGB events (eg, Lesbian Film Festival),^18,21 establishing political ties with LGB leaders, and supporting certain LGB causes, such as HIV/AIDS research.^18,20 Fallin et al. found that tobacco industry advertising receptivity was highest among bisexual females compared with other sexual minorities and heterosexual males and females,⁶ which may help to explain within group differences observed in our study. However, there is no research to date on exposure and receptivity to advertising for menthol cigarettes specifically among this population. There is a need for research to understand how tobacco industry marketing and other potential factors drive menthol use among sexual minorities, particularly among females.

This study has a few limitations. An analysis of missing data of key study variables (data not shown) found that those with complete data for sexual identity reported greater past 30-day smoking prevalence and lower preference for menthol cigarettes, compared with those who had missing data for sexual identity. Additionally, those with missing data for menthol preference (those who intentionally skipped the question, not those who did not get asked the question because they had not used cigarettes in the past 30 days) were younger, more likely to identify as Hispanic or non-Hispanic Other, and higher income. Based on these differences, missing values for sexual identity and menthol preference may have biased study findings. However, missing data are a common challenge in survey research and cannot be avoided in secondary data analysis. Another limitation is that these data are cross-sectional and therefore individuals could not be followed over time. Research is needed to understand the impact of menthol on cigarette smoking trajectories among sexual minorities, including initiation, progression from experimentation to more established patterns of use and cessation. Lastly, NSDUH relies on self-report to assess cigarette smoking, which can result in social desirability bias and recall bias.²² However, self-report of smoking behavior is standard in large-scale surveys and has found to be valid in prior research.²³

Overall, our study findings build upon the existing literature, which shows elevated menthol cigarette smoking among sexual minorities, by identifying sexual minority females as a particularly high-risk group when it comes to menthol cigarette smoking, even compared with other sexual minorities. Studies show that sexual minority females who smoke, initiate smoking^1,6 and progress to daily smoking¹ earlier than heterosexual females who smoke. Additionally, a study using data from NATS found that among those who smoke menthol cigarettes, identifying as a sexual minority was associated with lower intentions to quit smoking, compared with heterosexual adults who smoke menthol cigarettes.²⁴ Given what is known about the impact of menthol cigarette smoking on initiation, nicotine dependence, and cessation outcomes, elevated menthol use among sexual minority females may be partly responsible for these differences observed in the literature in initiation, progression to daily smoking and intentions to quit. More research is needed to understand the impact of menthol on patterns of cigarette smoking among sexual minorities, with a focus on understanding within group differences, in order to inform targeted interventions to prevent menthol cigarette initiation among sexual minorities, as well as promote smoking cessation among sexual minorities who smoke menthol cigarettes.

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.

Acknowledgments

The content is solely the responsibility of the authors and does not necessarily represent the official views of NCI or FDA.

Funding

This work was supported in part by National Cancer Institute and U.S. Food and Drug Administration Center for Tobacco Products (CTP) under U54CA229973 and the Rutgers Cancer Institute of New Jersey under P30CA07270-5931.

Declaration of Interests

The authors have no conflicts of interest to report.