Appendix D
Cross-Sectional Patterns and Longitudinal Transitions of Cigar Use by Type in the PATH Study
Authored by:1 Jihyoun Jeon2 and Yoonseo Mok,3 September 2021
INTRODUCTION
While premium cigar use is relatively common in the United States, studies of its prevalence and patterns are limited. The little evidence available comes from analyses of the National Adult Tobacco Survey and the Population Assessment of Tobacco and Health (PATH) study (Corey et al., 2014, 2018). In particular, Corey et al. (2018) analyzed adult cigar smoking patterns by cigar type, including premium cigars, and sociodemographic factors in PATH Wave 1 (2013–2014). They found an overall adult past 30-day prevalence of premium cigar smoking in 2013–2014 of 0.7 percent (95 percent confidence interval [CI]: 0.6–0.7) with use more common in men, non-Hispanic white people, people with some college education or more, and people with incomes of 200 percent of the federal poverty line or more.
Corey et al. also analyzed cigar use characteristics, such as the number of days smoked in the past 30 days, the number of cigars or cigarettes used per day, age at first regular use, duration of use, and concurrent use of cigars and cigarettes. They found that compared with users of other cigar products, premium cigar users smoke fewer days in the past 30 days (median: 1.7; interquartile range [IQR]: 0.0–4.8), smoke fewer cigars per
___________________
1 Committee member Rafael Meza oversaw this analysis, reviewed the results for accuracy, provided regular feedback to the authors, and edited the final paper.
2 Associate Research Scientist, Department of Epidemiology, University of Michigan.
3 Research Area Specialist, Department of Epidemiology, University of Michigan.
day (median: 0.1; IQR: 0.0–0.2), and have a lower prevalence of concurrent cigarette smoking (29.9 percent; 95 percent CI: 25.5–34.3 percent). However, as this was a cross-sectional study, no information was available on either rate of initiation or cessation of premium cigar or other cigar use or about transition rates between use of cigar types and cigarettes. Using PATH Waves 1–3, some studies have evaluated longitudinal patterns, such as initiation, cessation, reuptake, and relapse of tobacco product use (cigarettes, electronic nicotine delivery systems [ENDS)], cigars, hookah and smokeless tobacco), although they did not look into differences by cigar type (Edwards et al., 2020; Kasza et al., 2020a,b).
In Corey et al. (2018), traditional cigar use was categorized as premium versus nonpremium based on the tobacco blends, components, manufacturing process, and other characteristics associated with the usual brand reported by survey respondents. For traditional large cigar users with no brand information, the cutoff for premium was those who reported paying more than $2 per cigar. Some brands classified as premium by Corey et al. (2018) have characterizing flavors, which would exclude them from the premium cigar category under some classifications (FDA, 2020).
To inform the committee about the patterns of use of premium and other cigars, we conducted analyses of PATH data from Waves 1–5 that extend Corey et al. (2018) in several ways. First, we replicated that analysis for Waves 1–5, providing annual estimates of premium cigar and other cigar type use prevalence and patterns of use for 2013–2014, 2014–2015, 2015–2016, 2016–2017, and 2018–2019. We then calculated longitudinal transitions between use of premium cigars and other cigar types and cigarettes, also differentiating premium cigar use by frequency (number of days used in the past 30 days). Lastly, we conducted analyses of tobacco dependence for different cigar use groups, including exclusive premium cigar users.
METHODS
Data4
PATH is a longitudinal study of the civilian, noninstitutionalized U.S. population age 12+ by the National Institutes of Health and the Food and Drug Administration beginning in 2013–2014. PATH participants answer periodically a series of detailed questions about tobacco product use using Audio Computer-Assisted Self-Interviewing. PATH collected data annually from Waves 1–4 but switched to biennially beginning in Wave 5.
___________________
4 All supplemental materials cited in this paper are available in the project’s public access file and upon request from the National Academies Public Access Records Office at PARO@nas.edu.
For this analysis, we used the PATH sample age 18+: Wave 1 (September 2013–December 2014; N = 32,320; weighted response rate [WRR]: 74.0 percent), Wave 2 (October 2014–October 2015; N = 28,362; WRR: 83.2 percent), Wave 3 (October 2015–October 2016; N = 28,148; WRR: 78.4 percent), Wave 4 (December 2016–January 2018; N = 27,757 for Wave 1 cohort and 6,065 for Wave 4 replenishment sample; WRR: 73.5 percent for Wave 1 cohort and 68.0 percent for Wave 4 replenishment sample), and Wave 5 (December 2018–November 2019; N = 32,687; WRR: 88.0 percent) (HHS, 2021). Analyses relied on the PATH restricted-use file.
Measures
We followed similar procedures as Corey et al. (2018), extending their Wave 1 estimations to Waves 2–5, and updating the estimations for Wave 1. We briefly describe the variables and methods of the current analysis.
A question whether an individual has ever seen or heard of cigars, cigarillos, or filtered cigar before this study was asked of Wave 1 and replenishment Wave 4 participants at their study entry. Another set of questions assessed whether an individual smoked each cigar type, even one or two puffs in past 30 days. This second set of questions were asked of all adult respondents (continuing and aged-up adult respondents, or new cohort adult respondents who have ever seen or heard of traditional cigars, cigarillos, or filtered cigars) in Waves 2–5. These questionnaires in PATH display first images of traditional cigars with the physical characteristics and list examples of popular brands to all adult respondents (“Traditional cigars contain tightly rolled tobacco that is wrapped in a tobacco leaf. Some common brands of cigars include Macanudo, Romeo y Julieta, and Arturo Fuente [added Cohiba on Wave 3], but there are many others”). Then the questionnaires display images of cigarillos and filtered cigars with a description (“Cigarillos and filtered cigars are smaller than traditional cigars. They are usually brown. Some are the same size as cigarettes, and some come with tips or filters. Some common brands are Black&Mild, Swisher Sweets, Dutch Masters, Phillies Blunts, Prime Time, and Winchester [Cheyenne listed instead on Wave 3]”). Participants who reported smoking cigars “with a filter (like a cigarette filter)” were assigned as filtered cigar users, and those who reported “with a plastic or wooden tip” or “without a tip or filter” were assigned as cigarillos users.
Current Established Cigar and Cigarette Users
Current established cigar users were defined as those individuals who reported ever smoking the specific type “fairly regularly” and currently smoke every day or some days. Traditional cigar users were further differentiated into premium versus nonpremium with the usual brand smoked
(Supplement A). Traditional cigar brands reported by PATH participants were classified independently by three expert coders (Ganz, Villanti, and Sterling).5 Coding was based on the following premium cigar characteristics identified by the National Academies of Sciences, Engineering, and Medicine (the National Academies) committee: (1) handmade, (2) filler at least 50 percent natural long-leaf filler tobacco, (3) wrapped in whole leaf (i.e., not reconstituted tobacco), (4) weighs at least 6 lbs per 1,000, (5) no filters/tips, and (6) no characterizing flavor other than tobacco. Coders were able to look at brand and vendor websites. The three coders achieved 84 percent agreement (brands n = 110). Any disagreements (n = 21) were discussed among the three coders and an additional cigar expert (Delnevo).6 Ultimately, the three coders and the additional expert could not determine whether the cigar should be categorized as premium or nonpremium for four brands: Marsh Wheeling, Isla Del Sol, Java, and Acid. The group decided to treat these brands as nonpremium. Only three brands were coded differently from Corey et al. (2018). Specifically, Corey et al. (2018) categorized Acid, Optimo, and Marsh Wheeling as premium cigars, whereas these brands were coded as nonpremium for this analysis. In particular, Acid was classified as nonpremium because these cigars come in flavors. These decisions were consistent for both commissioned patterns of use studies. For individuals with missing usual brand information, those who reported paying ≥$2 per cigar were classified as premium cigar users, while those who reported paying <$2 per cigar were nonpremium. The analysis considered four cigar types: traditional premium cigars, traditional nonpremium cigars, cigarillos, and filtered cigars.
Current established cigarette users were defined as individuals who reported smoking at least 100 manufactured or roll-your-own cigarettes in their lifetime and now smoke cigarettes every day or some days.
Cigar Smoking Patterns
The number of cigars smoked in lifetime was stratified into three categories: 10 or fewer, 11–50, and 51+. Information for the number of days smoked in the past 30 days was collected for some days smokers, with every day smokers assumed to smoke on all 30 days. Number of cigars smoked per day was calculated for both every day and some days smokers by multiplying the number of days smoked in the past 30 days by the
___________________
5 Ganz is an author of a commissioned paper for this report (see Appendix C); Villanti and Sterling are members of the National Academies committee.
6 Delnevo oversaw the development of the National Survey on Drug Use and Health analysis (see Appendix C) and is a member of the National Academies committee.
average number of cigars smoked per day on such days divided by 30. Duration of cigar smoking for current established smokers was calculated by subtracting age at first regular use from current age. Current use of ≥1 other noncigar, noncigarette products was defined as having ever used “fairly regularly” at least one of ENDS, pipe tobacco, hookah, smokeless tobacco, snus, or dissolvable tobacco and now using it every day or some days (Note: dissolvable tobacco was excluded in Waves 3–5).
Cigar Purchasing
Cigar users reported whether they had a regular brand, the name of the brand of regular use or usually/last smoked, along with whether it was flavored or mentholated. They also specified how they purchase cigars (in person, from the Internet, by telephone, or do not buy their own cigars), where (cigar bar, convenience store/gas station, smoke shop/tobacco specialty or outlet store, or somewhere else), and their usual purchase size (single stick or box/pack). Price per cigar was calculated as the reported usual price paid divided by the number of cigars in the usual purchase size unit. Similar measures were also reported and calculated for cigarette users.
Reasons for Cigar Smoking
Total of 12 (Waves 1 and 2), 8 (Wave 3), 7 (Waves 4 and 5) reasons or beliefs for cigar smoking were asked to those reporting use (yes/no). The list of all items in each wave is available in Supplemental Table B5. Reasons that compared cigars with cigarettes were stratified by the participant’s cigarette smoking status (current, former, never).
Demographic Characteristics
Participants reported their demographic characteristics including sex (male, female), age in years (18–24, 25–34, 35–54, or 55+), race/ethnicity (non-Hispanic white, non-Hispanic Black/African American, non-Hispanic other/multi-race, or Hispanic), education (less than high school diploma, GED, high school diploma, some college/associate’s degree, completed college or more). Poverty status was assigned based on annual household income and household size as <100 percent of the federal poverty level (FPL), 100–<200 percent FPL, and ≥200 percent FPL. Information for poverty status was available only in Wave 1.
Cross-Sectional Analyses by Wave
Prevalence of cigar smoking by type (premium, nonpremium, cigarillos, and filtered cigars) and cigarette smoking were calculated using the “survey” package in the R statistical software version 4.1.1. Prevalence was calculated overall and according to sociodemographic characteristics, tobacco use patterns, purchasing behaviors, and reasons for use. Adjusted prevalence ratios (APRs) were calculated using a survey-weighted generalized linear model (“svyglm” function in R) with logit link function to examine associations between dual cigar and cigarette smoking versus cigar-only smoking, accounting for demographics and cigar use behaviors (daily versus nondaily smoking). To account for the complex survey design of PATH, all analyses used replicate weights and balanced repeated replication methods, Fay’s method with a factor ε = 0.3 as recommended by PATH (HHS, 2021). Prevalence with denominator <50 observations or a relative standard error of >30 percent was suppressed. Missing values for >5 percent of all eligible responses were treated as a separate category (e.g., poverty status); otherwise observations with missing values were dropped.
For cross-sectional (per wave) analyses, we estimated weighted prevalence and CIs using cross-sectional single-wave weights and their corresponding 100 replicate weights in each wave.
Longitudinal Transition Analyses
Transition Analysis of Cigar and Cigarette Use
We calculated longitudinal transition rates (percentage of users transitioning from one use state to another) for each PATH wave pair, using cross-sectional weights at the end wave of each pair. For example, the analysis for Wave 1 to 4 looked at the transition rates between Wave 1 and 4 in people who participated in both waves, regardless of their participation in other waves. In this specific analysis, the cross-sectional Wave 4 weights were used in the transition rate calculations. We first considered the following nine cigar and cigarette use states: (1) never cigar and cigarette use, (2) noncurrent cigar and cigarette use, (3) exclusive current established premium cigar use, (4) exclusive current established use of other cigar types (nonpremium cigars, cigarillos, or filtered cigars), (5) exclusive current established cigarette use, (6) dual current established use of premium cigars and other cigar types, (7) dual current established use of premium cigars and cigarettes, (8) dual current established use of other cigar types and cigarettes, and (9) poly tobacco use (i.e., current established use of premium cigars, other cigar types, and cigarettes). We also considered an alternative classification distinguishing premium cigar
users by frequency (seven use states): (1) never cigar and cigarette use, (2) noncurrent cigar and cigarette use, (3) exclusive current established premium cigar use with <6 days of use in the past 30 days, (4) exclusive current established premium cigar use with 6+ days of use in the past 30 days, (5) exclusive current established other combustible tobacco use (nonpremium cigars, cigarillos, filtered cigars, or cigarettes), (6) dual current established use of premium cigar with <6 days of use in the past 30 days and other combustible tobacco use, and (7) dual current established use of premium cigar with 6+ days of use in the past 30 days and other combustible tobacco use. Use of other tobacco products was not considered in these classifications.
For this longitudinal transition analysis, the sample consisted of all individuals who participated in Waves 1–5 (N = 18,925). Transition estimates were calculated using the all-wave weights for the Wave 1 cohort and the corresponding 100 replicate weights. We estimated the transition rates for all adults (ages 18+) and also stratified by age (18–34 and ages 35+).
Transition Analysis of Premium Cigar Users
To better assess the trajectories of premium cigar users, we conducted a second transition analysis restricting the sample to those who reported smoking premium cigar fairly regularly in at least one wave (N= 844). This analysis considered the same use states as in the previous analyses, with the addition of a “not in sample” state for individuals lost to followup or who did not participate in some waves. We used each individual baseline sample weight and the corresponding 100 replicate weights (i.e., the cross-sectional weights from the first wave in which the individual participated). We estimated the transition rates for all adults (ages 18+) and also stratified by age (ages 18–34 and 35+). As a sensitivity analysis, we also conducted the longitudinal trajectory analysis by limiting to Wave 1 cohort individuals who participated in all Waves 1–5 (N = 517) and who reported smoking premium cigars fairly regularly in at least one wave, using Wave 5 all-waves weights.
Tobacco Dependence
To compare the level of nicotine dependence of current established tobacco product users in PATH, we constructed a tobacco dependence score (TDS) based on the 16 items suggested by Strong et al. (2017): the Wisconsin Inventory of Smoking Dependence Motives or WISDM (11 items), Nicotine Dependence Syndrome Scale or NDSS (4 items), and Diagnostic and Statistical Manual criteria (1 items). The nicotine dependence domains of these items consist of “Automaticity,” “Craving,” “Loss
of Control,” “Tolerance,” “Negative Reinforcement,” “Cognitive Enhancement,” “Affiliative Attachment,” and “Withdrawal.” For the 15 items with five response categories in order of increasing symptom level (i.e., 1 being “no symptom” and 5 being “extremely true”), we recoded them as a three-category scale, placing 1, 2–3, and 4–5 into the categories 0, 1, and 2, respectively. We then rescaled the three-level items by multiplying them by 50, i.e., (0,1,2) × 50 = (0,50,100). For the one item with binary response, we coded 0 for “No” and 1 for “Yes,” then rescaled it by multiplying it by 100, i.e., (0,1) × 100 = (0,100). We calculated a TDS per individual per wave as the linear average of these 16 responses. We then compared the mean and median TDS for the following exclusive tobacco product use categories: premium cigars, nonpremium cigars, cigarillos, filtered cigars, cigarettes, smokeless tobacco, and hookah. For each group, we also calculated the percentage of users reporting at least one dependence symptom, the difference in mean TDS between daily and nondaily users of each product, the difference in mean TDS between those who used it less than 6 days and 6+ days in the past 30 days, and the linear association of the number of days used in the past 30 days with TDS. Note that for tobacco dependence, some respondents might answer these questions with a specific product in mind, but if they are dual or poly users (even if that use is experimental and not established by the PATH study definition), then they could be answering these questions referring to their different tobacco products.
RESULTS
Cross-Sectional Patterns of Use
This study analyzed use patterns of four cigar types (premium, nonpremium, cigarillos, and filtered cigars) and cigarettes among U.S. adults across different periods from 2013 to 2019. For the purpose of illustration, we describe the results for PATH Wave 4 (2016–2017) in the main text (see Tables D-1–D-4). PATH Wave 4 added a new sample of users (replenishment sample) to address loss to follow-up in previous waves and produce a truly nationally representative sample as in Wave 1. Results for PATH Waves 1–5 are available in Supplemental Tables B1–B4. The analyses of reasons for cigar product use for Waves 1–5 are shown in Supplemental Table B5. Results using cigar use definitions without the fairly regularly restriction or the 100 cigarettes in their lifetime criteria for cigarette smoking are presented in Supplemental Table B6.
Demographic Characteristics of Cigar and Cigarette Users
The overall adult prevalence of current established adult tobacco use in PATH Wave 4 (2016–2017) was 0.7 percent (95 percent CI: 0.6–0.8 percent) for premium cigars, 0.5 percent (95 percent CI: 0.4 –0.5 percent) for nonpremium cigars, 1.5 percent (95 percent CI: 1.4–1.6 percent) for cigarillos, 0.8 percent (95 percent CI: 0.8–0.9 percent) for filtered cigars, and 17.7 percent (95 percent CI: 17.3–18.1 percent) for cigarettes (see Table D-1). The absolute use prevalence for each cigar type varies slightly by wave, but the general use patterns by sociodemographic groups for all cigar types are consistent across waves (see Figure D-1 and Supplemental Table B1). The majority of adult cigar users were male (70.2–97.7 percent). In contrast, 53.9 percent (95 percent CI: 52.8–55.1 percent) of cigarette users were male and 46.1 percent (95 percent CI: 44.9–47.2 percent) were female. Cigarillos were smoked by young adults aged 18–34 years at relatively higher rates compared to other cigar types and cigarettes (54.8 percent versus 37.0 percent for premium cigar, 28.6 percent for nonpremium cigar, 31.5 percent for filtered cigar, and 34.4 percent for cigarette users). The use prevalence among non-Hispanic Black adults was higher for cigarillos (34.1 percent; 95 percent CI: 29.2–39.4 percent) compared to premium cigars (6.2 percent; 95 percent CI: 3.4–11.1 percent), and other cigar types and cigarettes (13.2–24.1 percent). The majority of the adult premium cigar users were non-Hispanic white people (79.5 percent; 95 percent CI: 73.5–84.5 percent). High school diploma, GED, or less comprised 52.5 percent of nonpremium cigar, 54.2 percent of cigarillo, 62.8 percent of filtered cigar, and 57.4 percent of cigarette users but only 22.9 percent of premium cigar users. Premium cigar users were predominantly male, non-Hispanic white people, with some college or more education. Results for other waves are presented in Supplemental Table B1.
Cigar and Cigarette Smoking Patterns
Cigar smoking patterns and use behaviors varied by type (see Table D-2). About half of the established premium, nonpremium cigar, and cigarillo users had smoked >50 cigars in their lifetime (46.1–51.8 percent), while filtered cigar users had similar distribution across the categories (only 35.3 percent smoked >50 filtered cigars in their lifetime). The prevalence of daily cigar smoking was highest for filtered cigar (39.6 percent; 95 percent CI: 35.2–44.3 percent) and lowest for premium cigars (5.2 percent; 95 percent CI: 2.2–11.4 percent). In contrast, the prevalence of daily use for cigarette users was 76.4 percent (95 percent CI: 75.4–77.4 percent).
The number of cigars smoked per day was greater for filtered cigars (median: 1.0 cigars/day; IQR: 0.0–8.3) compared with the other cigar types
(median: 0.1–0.3 cigars/day), and lowest for premium cigars (median: 0.1 cigars/day, IQR: 0.0–0.2). For cigarette users, the number of cigarettes smoked per day had a median of 10.0 (IQR: 4.7–20.0). Age at first regular use was older for filtered cigars (median: 30.0 years) compared to the other cigar types (median range 19–25 years) and cigarettes (median: 17.0 years). The current use percentage of one or more other cigar types among cigar users was highest in nonpremium smokers (52.8 percent; 95 percent CI: 44.5–60.8 percent) and lowest in premium smokers (16.4 percent; 95 percent CI: 11.2–23.3 percent). This was lower for cigarette users (8.2 percent; 95 percent CI: 7.7–8.7 percent). The usage of one or more noncigar/noncigarette products was similar across users of all cigar types but lower for cigarette users. Concurrent cigarette smoking was lowest among premium cigar users (25.7 percent; 95 percent CI: 19.3–33.3 percent), with a prevalence of 50.1–70.5 percent for the other cigar type users. These patterns were consistent for the other PATH waves, except for the proportion of cigar users who smoked >50 cigars in their lifetime, which decreased over time, especially for premium and nonpremium cigar users (Supplemental Table B2).
Tobacco Product Characteristics and Purchasing Behaviors
The majority of nonpremium cigar, cigarillo, filtered cigar, and cigarette users reported having a regular brand (68.0–92.5 percent), while only 38.4 percent of premium cigar users did so (Table D-3). Black & Mild was the leading brand of nonpremium cigars (23.6 percent; 95 percent CI: 16.0–31.2 percent), cigarillos (51.0 percent; 95 percent CI: 46.8–55.2 percent), and filtered cigars (20.7 percent; 95 percent CI: 14.8–26.7 percent). Swisher Sweets was reported as the second most commonly used brand among cigarillo users (21.1 percent; 95 percent CI: 16.9–25.4 percent). Acid, which was classified here as nonpremium, was the second most commonly used brand among nonpremium cigar users (12.7 percent; 95 percent CI: 8.0–17.5 percent). Cohiba (20.3 percent; 95 percent CI: 14.0–26.4 percent) and Arturo Fuente (17.9 percent; 95 percent CI: 11.4–24.5 percent) were the two most common brands reported by premium cigar users. Marlboro, Newport, Camel, Pall Mall, and American Spirit were the most common brands reported by cigarette users. Flavored or mentholated regular brands were less likely for users of premium cigars (7.6 percent; 95 percent CI: 5.5–10.5 percent) compared to other cigar types (48.4–55.7 percent). In comparison, 38.4 percent (95 percent CI: 36.8–40.0 percent) of cigarette users reported using a mentholated brand. The majority of nonpremium cigar, cigarillo, filtered cigar, and cigarette users purchased in person (84.0–94.9 percent), and most of them bought in convenience store/gas stations (64.5–88.1 percent). In contrast, about a quarter of pre-
mium cigar users did not buy in person and mostly bought in smoke shop/tobacco specialty or outlet stores (44.3 percent; 95 percent CI: 35.2–53.8 percent) or cigar bars (35.4 percent; 95 percent CI: 28.4–43.1 percent). The median price paid per stick was lowest for filtered cigars ($0.11; IQR: $0.07–$0.26) and highest for premium cigars ($7.00; IQR: $4.00–$10.00). The median price paid per stick for cigarette users was higher than that of filtered cigars, but lower than that of other cigar products ($0.30; IQR: $0.25–$0.35). Although the most popular cigar brands varied over time, the patterns of purchasing were consistent throughout all PATH waves (Supplemental Table B3).
Factors Associated with Dual Cigar and Cigarette Smoking Versus Cigar-Only Smoking
Among current established cigar users, those smoking cigarillos or filtered cigars were more likely to be dual cigar and cigarette smokers (APR: 1.37; 95 percent CI: 1.17–1.59 for cigarillo users; APR: 1.58; 95 percent CI: 1.41–1.77 for filtered cigar users). On the other hand, those smoking premium or nonpremium cigars were less likely to be dual users (APR: 0.53; 95 percent CI: 0.41–0.69 for premium cigar users; APR: 0.80; 95 percent CI: 0.67–0.95 for nonpremium cigar users) (see Table D-4). Cigar users who use other tobacco products (ENDS, pipe tobacco, hookah, smokeless tobacco, or snus) were more likely to smoke cigarettes (APR: 1.14; 95 percent CI: 1.04–1.25). Non-Hispanic Black cigar users were less likely to smoke cigarettes compared to non-Hispanic white people (APR: 0.85; 95 percent CI: 0.74–0.97). Cigar users with education level of GED, high school diploma, or less were more likely to be dual cigar and cigarette users than those who had some college/associate degree or more (APR: 1.38; 95 percent CI: 1.26–1.51). Daily cigar users were less likely to smoke cigarettes compared to nondaily cigar users (APR: 0.82; 95 percent CI: 0.72–0.92). These patterns were consistent across all PATH waves (Supplemental Table B4).
Reasons for Using Cigar Products
While the reasons for cigar use varied by cigar type and cigarette smoking status, half or more of cigar users reported either “socializing while smoking” or “availability of products in favorite flavors” (see Supplemental Table B5). Affordability was another common reason for use reported by users of nonpremium cigars (range across waves 57.1–64.0 percent), cigarillos (range across waves 68.0–73.0 percent), and filtered cigars (range across waves 73.8–80.2 percent) but not premium cigars (range across waves 21.6–30.2 percent). About half of filtered cigar users
overall (range across waves 46.9–52.9 percent) indicated that they feel like smoking a regular cigarette while they smoke a filtered cigar; a smaller percentage of users of the other cigar types reported this as a reason for use (range across waves 2.7–6.7 percent in premium cigar, 22.6–25.8 percent in nonpremium cigar, and 24.9–28.6 percent in cigarillo users).
Longitudinal Transition Analyses
Transition Analysis of Cigar and Cigarette Use
Cross-sectional unweighted counts, weighted prevalence, and 95 percent CIs of cigar and cigarette use were estimated for adults (ages 18+) in PATH Waves 1–5. Stratified counts by age were also estimated (18–34 versus 35+). These estimates are shown in Supplement C1 (9 categories of tobacco use state) and Supplement C2 (7 categories of tobacco use state).
We present empirical weighted transition probabilities of tobacco use state between any wave pairs in 1-year, 2-year, 3-year, 4-year, or 5-year intervals (see Supplements C1 and C2). For example, 1-year transition probabilities were estimated from Wave 1 to 2, Wave 2 to 3, and Wave 3 to 4 and 5-year transition probabilities from Wave 1 to 5. Note that PATH collected data annually for Waves 1–4 but switched to biennially beginning in Wave 5; therefore, there is a 2-year gap between Waves 4 and 5. We also calculated average 1-year (1-wave) transitions across Waves 1 to 4. Figure D-2 shows a heatmap of the 1-year average transition probabilities for all adults (18+). About 75 percent of exclusive premium cigar users kept smoking premium cigars (sum of 69.2 percent as exclusive premium cigar users, 2.5 and 2.8 percent as dual users with other cigar types and cigarettes, respectively, and 0.5 percent as poly tobacco users) in the following year (see Figure D-2a). Most exclusive premium cigar users transitioning away do so to noncurrent use (18.8 percent) with only a small fraction transitioning to other products (6.1 percent). About 59 percent of exclusive other cigar type (nonpremium cigars, cigarillos, or filtered cigars) users kept smoking cigars as exclusive users (46.9 percent) or dual users with premium cigars (2.2 percent) or cigarettes (9.6 percent), or poly tobacco users (0.3 percent) in the following year. More than a quarter of exclusive other cigar type users discontinued use within a year (32.1 percent). Dual use of premium cigars with either other cigar types or cigarettes was relatively transient, with only slightly more than 40 percent of individuals staying as dual users in the following year. In particular, about 34 percent of dual users of premium cigars and cigarettes became exclusive cigarette users within a year.
About 68.2 percent of exclusive premium cigar users who smoke less than 6 days in the past 30 days continued smoking premium cigars in the following year as either exclusive users (63.0 percent) or dual users
with other combustible tobacco products (5.2 percent) (see Figure D-2b). While only 4 percent of less frequent exclusive premium cigar users (i.e., used less than 6 days in the past 30 days) increased use frequency within a year, about 20 percent of more frequent exclusive premium cigar users (i.e., 6+ days in the past 30 days) decreased use frequency. Less frequent exclusive premium cigar users were more likely to discontinue use within a year compared to more frequent users (22.7 versus 4.3 percent). Dual premium cigar use with other combustible tobacco products is relatively transient; only 44.5 percent of less frequent and 15.8 percent of more frequent premium cigar users remained as dual users. About 35.2 percent of less frequent and 40.9 percent of more frequent dual premium cigar users discontinued premium cigar use and became exclusive other combustible tobacco product users within a year.
Figure D-3 shows longitudinal trajectories (alluvial plot) of adult exclusive premium cigar users at baseline (Wave 1). Slightly more than half remained as exclusive premium cigar users in all waves. About 35 percent of them discontinued use, and 7.2 percent switched to other tobacco products by Wave 5. Similar alluvial plots of longitudinal trajectories for other use categories at baseline and analyses stratified by age (18–34 versus 35+) are shown in Supplement C1. Alluvial plots for longitudinal trajectories based on the seven state categories differentiating frequency of premium cigar use are shown in Supplement C2.
Transition Analysis of Premium Cigar Use at Any Wave
Figure D-4 shows longitudinal trajectories of cigar and cigarette use for those who smoked premium cigars fairly regularly in at least one wave. About 14 percent of adults included in this analysis were not present at Wave 1 but entered the study in following years (either aged up from the youth sample or included in the replenishment sample in Wave 4). Participants dropping from the study in the following years were categorized as “Not in sample” (28.8 percent in Wave 5). The distribution of tobacco use categories remained somewhat consistent across waves, with the exception of dual premium cigar users with cigarettes, whose prevalence kept decreasing (9.1 percent in Wave 1, 7.6 percent in Wave 2, 7.3 percent in Wave 3, 6.4 percent in Wave 4, and 5.1 percent in Wave 5). The overall percentage of premium cigar users (exclusive, dual, or poly users) remained similar across waves (42.7 percent in Wave 1, 36.2 percent in Wave 2, 42.1 percent in Wave 3, 40.1 percent in Wave 4, and 40.3 percent in Wave 5). We also provide the alluvial plot for longitudinal trajectory by limiting to the Wave 1 cohort that participated in Waves 1–5 and reported smoking premium cigar fairly regularly in at least one wave (see Supplement D1). In this sensitivity analysis, the overall percentage of premium cigar users was higher in recent waves (41.9 percent in Wave
1, 45.7 percent in Wave 2, 51.1 percent in Wave 3, 49.0 percent in Wave 4, and 53.8 percent in Wave 5). Supplements D1 and D2 present alluvial plots showing longitudinal trajectories of ever premium cigar users in alternative tobacco use categories (seven use states) and also stratified by age (18–34 versus 35+).
Tobacco Dependence among Exclusive Tobacco Users
Table D-5 shows the mean (95 percent CI) and median (IQR) of TDS estimated for each exclusive tobacco product use group: premium cigars, nonpremium cigars, cigarillos, filtered cigars, cigarettes, smokeless tobacco, hookah, and ENDS. Exclusive cigarette (mean: 51.1 and 95 percent CI: 50.6–51.5; median: 53.1; IQR: 28.1–75.0 in Wave 4) and smokeless tobacco (mean: 44.0 and 95 percent CI: 41.9–46.0; median: 43.8, IQR: 18.8–65.6 in Wave 4) users have the highest levels of TDS. Exclusive premium cigar (mean: 6.6 and 95 percent CI: 5.4–7.8; median: 0.0, IQR: 0.0–6.3 in Wave 4) and hookah (mean: 6.8 and 95 percent CI: 5.1–8.6; median: 0.0, IQR: 0.0–6.3 in Wave 4) users have the lowest levels of TDS. The mean and median TDS in the other exclusive tobacco product (nonpremium cigars, cigarillos, filtered cigars, and ENDS) users were 10.4–36.7 and 0.0–34.7, respectively. More than half of exclusive users of premium cigars and hookah did not report any dependence symptoms in Waves 3–5.
The difference in the level of tobacco dependence between nondaily and daily exclusive users was significant in all tobacco groups, except for premium cigar users in Waves 4 and 5. In Wave 4, the difference in mean TDS between nondaily and daily exclusive users was 13.5 (95 percent CI: 4.7–31.6) for premium cigars, 20.8 (95 percent CI: 8.0–33.6) for nonpremium cigars, 24.1 (95 percent CI: 18.9–29.3) for cigarillos, 17.1 (95 percent CI: 4.4–29.9) for filtered cigars, 36.6 (95 percent CI: 35.7–37.5) for cigarettes, 27.6 (95 percent CI: 23.8–31.4) for smokeless tobacco, 40.4 (95 percent CI: 34.3–46.5) for hookah, and 29.4 (95 percent CI: 26.0–32.8) for ENDS.
Similarly, the difference in the level of tobacco dependence between less frequent (less than 6 days in the past 30 days) and more frequent (6+ days in the past 30 days) exclusive users was significant in all tobacco groups, except for filtered cigars in Wave 4. For example, in Wave 4, the difference in mean TDS between less and more frequent exclusive users was 12.6 (95 percent CI: 4.0–21.2) for premium cigars, 15.7 (95 percent CI: 7.0–24.3) for nonpremium cigars, 17.1 (95 percent CI: 13.0–21.1) for cigarillos, 13.9 (95 percent CI: 2.6–30.5) for filtered cigars, 39.2 (95 percent CI: 37.6–40.7) for cigarettes, 29.6 (95 percent CI: 23.9–35.4) for smokeless tobacco, 13.8 (95 percent CI: 1.9–25.7) for hookah, and 30.2 (95 percent CI: 26.3–34.2) for ENDS.
Significant linear associations between the use frequency in the past 30 days and level of tobacco dependence were observed in all exclusive tobacco users, except for hookah users in Wave 4. Mean and median of TDS estimates for premium cigar users differentiated by frequency (less than 6 versus 6+ days used in the past 30 days) are presented in Supplemental Table B7.
LIMITATIONS
The main limitation of the study is that the designation of premium versus nonpremium traditional cigar use was based on the usual brand or price reported and not on direct assessment by study participants. While this approach was validated by Corey et al. (2018) and three experienced coders performed the brand classification, some traditional cigar users might have been misclassified. Like Corey et al. (2018), we did not adjust estimates of the number of cigars smoked by size or weight, precluding comparisons of exposure dose and amount smoked. Another important limitation is the lack of information in PATH about the cigar smokers’ level of inhalation, which precluded characterization of inhalation patterns and analyses of dependence by inhalation. Inhalation has been shown to be an important determinant of health effects of cigar smoking (NCI, 1998), so understanding these behaviors is critical for health effects risk assessment. One more limitation is the relatively short period of analysis. While PATH allows for cross-sectional trend and longitudinal analyses of cigar use, it covers only 6 years (2013–2019), precluding analyses of long-term trends and patterns. Finally, while we evaluated patterns of use of different cigar types and cigarettes, we did not do so for other tobacco products, such as smokeless tobacco, e-cigarettes and other ENDS, hookah, or pipe. However, the analysis did evaluate the proportion of cigar users that use cigarettes or other tobacco products at each PATH wave.
SUMMARY OF KEY PATH RESULTS
- Premium cigars are predominantly consumed by male smokers, non-Hispanic white people, and individuals with some college or more education.
- The proportion of cigar users who smoked >50 cigars in their lifetime decreased over time, especially among premium and nonpremium cigar users.
- The majority of premium cigar users were nondaily cigar smokers.
- Premium cigar users were considerably less likely to smoke cigarettes or other cigar types in comparison with other cigar type users.
- In all waves, while other type cigars and cigarettes were purchased in convenience store/gas stations, premium cigars were purchased in either cigar bars or smoke shops/tobacco specialty or outlet stores. Premium cigar users were less likely to use flavored or mentholated brands compared to users of the other cigar types or cigarettes.
- Dual use of cigar and cigarettes was less likely among non-Hispanic Black people compared to non-Hispanic white people, daily smokers compared to nondaily smokers, and those who have some college/associate degree or more compared to those with education level of GED, high school diploma, or less.
- Two common reasons reported for smoking cigars were “socializing while smoking” or “availability of products in favorite flavors.” Affordability was another common reason for smoking nonpremium cigars, cigarillos, and filtered cigars, but not for premium cigars.
- About three-quarters of exclusive premium cigar smokers remained so in the following year. Dual use of premium cigars with either other cigar types or cigarettes is somewhat transient.
- Exclusive premium cigar users who smoked less than 6 days in the past 30 days were more likely to discontinue use within a year compared to those who smoked 6+ days in the past 30 days. About 20 percent of 6+ days exclusive premium cigar users became <6 days users within a year.
- Among exclusive premium cigar users in Wave 1, slightly more than half remained as exclusive premium cigar users in Wave 5 and about 35 percent discontinued use by Wave 5.
- Among those who smoked premium cigars fairly regularly in at least one wave, the overall percentage of premium cigar users (exclusive, dual, or poly users) remained similar across waves, but the prevalence of dual premium cigar users with cigarettes decreased over time.
- The level of tobacco dependence is lowest in exclusive premium cigar and hookah users.
- The difference in the level of tobacco dependence in nondaily versus daily (or less versus more frequent) exclusive users is significant for all tobacco products, except for daily versus nondaily premium cigar users in Waves 4 and 5 and less versus more frequent filtered cigar users in Wave 4.
REFERENCES
Corey, C. G., B. A. King, B. N. Coleman, C. D. Delnevo, C. G. Husten, B. K. Ambrose, and B. J. Apelberg. 2014. Little filtered cigar, cigarillo, and premium cigar smoking among adults—United States, 2012–2013. Morbidity and Mortality Weekly Report 63(30):650–654.
Corey, C. G., E. Holder-Hayes, A. B. Nguyen, C. D. Delnevo, B. L. Rostron, M. Bansal-Travers, H. L. Kimmel, A. Koblitz, E. Lambert, J. L. Pearson, E. Sharma, C. Tworek, A. J. Hyland, K. P. Conway, B. K. Ambrose, and N. Borek. 2018. U.S. adult cigar smoking patterns, purchasing behaviors, and reasons for use according to cigar type: Findings from the Population Assessment of Tobacco and Health (PATH) study, 2013–2014. Nicotine & Tobacco Research 20(12):1457–1466.
Edwards, K. C., E. Sharma, M. J. Halenar, K. A. Taylor, K. A. Kasza, H. Day, H. T. Hammad, G. Anic, M. Bansal-Travers, J. Limpert, L. D. Gardner, N. Borek, H. L. Kimmel, W. M. Compton, A. Hyland, and C. A. Stanton. 2020. Longitudinal pathways of exclusive and polytobacco cigar use among youth, young adults and adults in the USA: Findings from the PATH study Waves 1–3 (2013–2016). Tobacco Control 29(Suppl 3):s163–s169.
FDA (Food and Drug Administration). 2020. CTP statement on premarket authorization requirements for premium cigars. https://www.fda.gov/tobacco-products/ctp-newsroom/ctp-statement-premarket-authorization-requirements-premium-cigars (accessed September 24, 2021).
HHS (Department of Health and Human Services). 2021. Population Assessment of Tobacco and Health (PATH) study [United States] restricted-use files. Inter-university consortium for political and social research [distributor], 2021-06-29. https://doi.org/10.3886/ICPSR36231.v27 (accessed November 11, 2021).
Kasza, K. A., K. C. Edwards, Z. Tang, C. A. Stanton, E. Sharma, M. J. Halenar, K. A. Taylor, E. Donaldson, L. C. Hull, H. Day, M. Bansal-Travers, J. Limpert, I. Zandberg, L. D. Gardner, H. T. Hammad, N. Borek, H. L. Kimmel, W. M. Compton, and A. Hyland. 2020a. Correlates of tobacco product initiation among youth and adults in the USA: Findings from the PATH study Waves 1–3 (2013–2016). Tobacco Control 29(Suppl 3):s191–s202.
Kasza, K. A., K. C. Edwards, Z. Tang, C. A. Stanton, E. Sharma, M. J. Halenar, K. A. Taylor, E. A. Donaldson, L. C. Hull, M. Bansal-Travers, J. Limpert, I. Zandberg, L. D. Gardner, H. T. Hammad, N. Borek, H. L. Kimmel, W. M. Compton, and A. Hyland. 2020b. Correlates of tobacco product cessation among youth and adults in the USA: Findings from the PATH study Waves 1–3 (2013–2016). Tobacco Control 29(Suppl 3):s203–s215.
NCI (National Cancer Institute). 1998. Cigars: Health effects and trends. Tobacco control monograph no. 9. NIH pub. No. 98-4302. Bethesda, MD: National Cancer Institute.
Strong, D. R., J. Pearson, S. Ehlke, T. Kirchner, D. Abrams, K. Taylor, W. M. Compton, K. P. Conway, E. Lambert, V. R. Green, L. C. Hull, S. E. Evans, K. M. Cummings, M. Goniewicz, A. Hyland, and R. Niaura. 2017. Indicators of dependence for different types of tobacco product users: Descriptive findings from Wave 1 (2013–2014) of the Population Assessment of Tobacco and Health (PATH) study. Drug and Alcohol Dependence 178:257–266.
TABLE D-1 Demographic Characteristics of Adult Current Established Traditional Cigar (Premium, Nonpremium), Cigarillo, Filtered Cigar, and Cigarette Smokers, PATH Study Wave 4, 2016–2017
| Premium Cigarsa (n = 338) | Nonpremium Cigarsa (n = 238) | Cigarillos (n = 992) | Filtered Cigars (n = 486) | Cigarettes (n = 9,915) | |
|---|---|---|---|---|---|
| % (95% CI) | % (95% CI) | % (95% CI) | % (95% CI) | % (95% CI) | |
| Overall adult prevalence | 0.7 (0.6–0.8) | 0.5 (0.4–0.5) | 1.5 (1.4–1.6) | 0.8 (0.8–0.9) | 17.7 (17.3–18.1) |
| Sex | |||||
| Male | 97.7 (95.2–98.9) | 86.6 (79.9–91.3) | 70.2 (67.2–73.0) | 71.7 (67.2–75.9) | 53.9 (52.8–55.1) |
| Female | 2.3 (1.1–4.8) | 13.4 (8.7–20.1) | 29.8 (27.0–32.8) | 28.3 (24.1–32.8) | 46.1 (44.9–47.2) |
| Age group (years) | |||||
| 18–24 | 9.7 (6.3–14.7) | 8.1 (4.9–13.1) | 21.7 (18.5–25.2) | 10.9 (8.2–14.5) | 10.1 (9.4–10.8) |
| 25–34 | 27.3 (21.4–34.3) | 20.5 (15.5–26.4) | 33.1 (28.9–37.5) | 20.6 (16.9–25.0) | 24.3 (23.2–25.6) |
| 35–54 | 28.2 (22.1–35.2) | 38.3 (29.7–47.8) | 33.0 (28.5–37.8) | 38.2 (32.8–43.9) | 38.9 (36.8–41.0) |
| 55+ | 34.8 (30.8–38.9) | 33.1 (25.8–41.4) | 12.2 (9.6–15.5) | 30.3 (24.8–36.3) | 26.6 (25.2–28.2) |
| Race/ethnicity | |||||
| White, non-Hispanic | 79.5 (73.5–84.5) | 60.9 (50.2–70.7) | 47.4 (41.4–53.6) | 57.6 (49.3–65.6) | 68.4 (67.2–69.6) |
| Black/AA, non-Hispanic | 6.2 (3.4–11.1) | 24.1 (15.4–35.7) | 34.1 (29.2–39.4) | 20.2 (14.7–27.0) | 13.2 (12.5–14.0) |
| Other or multi-race, non-Hispanic | 5.0 (2.8–8.9) | 4.0 (2.3–6.8) | 4.6 (3.3–6.3) | 5.2 (3.6–7.5) | 5.6 (5.0–6.3) |
| Hispanic | 9.2 (5.9–14.2) | 11.1 (7.7–15.7) | 13.9 (11.0–17.3) | 17.0 (13.2–21.7) | 12.7 (11.7–13.9) |
| Education | |||||
| Less than high school diploma | 4.0 (2.0–7.5) | 14.2 (9.2–21.4) | 17.0 (14.8–19.4) | 23.0 (18.8–27.8) | 16.3 (15.3–17.3) |
| GED | 4.6 (2.6–8.2) | 7.9 (4.7–13.0) | 9.2 (6.7–12.4) | 12.0 (8.9–16.1) | 11.5 (10.4–12.7) |
| High school diploma | 14.3 (9.0–22.0) | 30.4 (23.3–38.5) | 28.0 (24.7–31.5) | 27.8 (23.3–32.8) | 29.6 (27.9–31.4) |
| Some college/associate degree | 30.8 (25.5–36.6) | 34.3 (27.1–42.4) | 37.0 (32.5–41.8) | 28.5 (25.8–31.3) | 32.0 (30.1–34.0) |
| Complete college or more | 46.3 (37.7–55.1) | 13.1 (7.5–22.1) | 8.9 (7.0–11.1) | 8.7 (5.6–13.2) | 10.5 (9.7–11.4) |
a Among traditional established cigar smokers, 4.8% (n = 29) could not be assigned as either a premium or nonpremium smoker after assessing responses to usual brand (Supplement A).
NOTE: AA = African American; CI = logit-transformed Wald-type confidence interval; PATH = Population Assessment of Tobacco and Health.
TABLE D-2 Smoking Patterns Among Adult Current Established Traditional Cigar (Premium, Nonpremium), Cigarillo, Filtered Cigar, and Cigarette Smokers, PATH Study Wave 4, 2016–2017
| Premium Cigars | Nonpremium Cigars | Cigarillos | Filtered Cigars | Cigarettesa | |
|---|---|---|---|---|---|
| % (95% CI) | % (95% CI) | % (95% CI) | % (95% CI) | % (95% CI) | |
| Lifetime cigars smoked | |||||
| <1–10 cigars | 5.5 (2.3–12.9) | 23.9 (17.7–31.4) | 14.0 (10.8–17.8) | 28.8 (21.3–37.7) | NA |
| 11–50 cigars | 47.9 (38.2–57.8) | 30.1 (19.3–43.6) | 34.2 (30.5–38.2) | 25.9 (18.6–34.8) | NA |
| 51+ cigars | 46.5 (36.5–56.8) | 46.1 (36.4–56.1) | 51.8 (47.6–55.9) | 35.3 (37.6–53.3) | NA |
| Now smoke product every day | 5.2 (2.2–11.4) | 22.3 (14.8–32.1) | 19.1 (16.7–21.8) | 39.6 (35.2–44.3) | 76.4 (75.4–77.4) |
| Days smoked in past 30 daysb (median, IQR) | 1.0 (0.0–4.0) | 5.0 (1.0–25.1) | 5.0 (1.0–25.0) | 19.3 (1.0–30.0) | 30.0 (30.0–30.0) |
| Number of cigars or cigarettes per dayc (median, IQR) | 0.1 (0.0–0.2) | 0.2 (0.0–1.0) | 0.3 (0.0–1.5) | 1.0 (0.0–8.3) | 10.0 (4.7–20.0) |
| Age (years) at first regular used (median, IQR) | 25.0 (20.0–31.0) | 22.0 (18.0–35.0) | 19.0 (16.0–25.0) | 30.0 (19.0–43.1) | 17.0 (15.0–19.0) |
| Duration (years) since first regular use (median, IQR) | 13.0 (6.0–25.0) | 18.0 (10.1–26.0) | 11.0 (6.0–17.0) | 9.0 (5.0–18.0) | 24.0 (14.0–38.0) |
| Currently use other cigar type(s)e,f | 16.4 (11.2–23.3) | 52.8 (44.5–60.8) | 32.8 (29.1–36.7) | 42.2 (36.8–47.9) | 8.2 (7.7–8.7) |
| Currently use noncigar, noncigarette product(s)g | 26.1 (21.0–31.9) | 28.8 (21.5–37.4) | 28.3 (24.2–32.9) | 26.8 (21.1–33.2) | 15.4 (14.1–16.8) |
| Cigarette smoking statush | |||||
| Current established smoker | 25.7 (19.3–33.3) | 50.1 (43.8–56.4) | 60.2 (57.2–63.2) | 70.5 (64.9–75.6) | NA |
| Former established smoker | 40.6 (33.1–48.6) | 25.9 (16.7–37.8) | 15.1 (11.9–18.9) | 12.4 (8.8–17.1) | NA |
| Never smoker | 33.7 (26.4–41.8) | 24.0 (14.7–36.8) | 24.7 (21.9–27.8) | 17.1 (13.1–22.0) | NA |
a When a respondent reported smoking both manufactured cigarettes and roll-your-own (RYO) cigarettes (n = 753), for certain topics, they were asked separate questions about each product. For dual manufactured cigarette and RYO smokers, the responses to manufactured cigarette products are provided; otherwise, responses reflect the single cigarette type the respondent reported smoking.
b Number of days using the product in past 30 days was asked of those who now smoke cigars some days; every day smokers were assumed to smoke on all 30 days.
c Respondents reporting smoking less than one cigar per day on the days smoked were assigned as smoking 0.5 cigars per day.
d Those reporting age at first regular use <6 years were assigned a value of 6 years.
e For current cigarette smokers, “currently use other cigar products” refers to current smoking of one or more cigar products.
f If respondent was missing status for one cigar product and did not smoke the other cigar product, then they were treated as not smoking other cigar types.
g Current use of noncigar, noncigarette product(s) defined as having ever used one or more of the following tobacco products “fairly regularly” and now using that product every or some days: ENDS, pipe tobacco, hookah, smokeless tobacco, or snus. If respondent reported not using any other tobacco product or some combination of not using and missing tobacco product use status, then they were treated as not using any noncigar, noncigarette products.
h Former established cigarette smokers had to have smoked at least 100 cigarettes in their lifetime and now smoke cigarettes not at all; never cigarette smokers had to smoke fewer than 100 cigarettes in their lifetime.
NOTE: CI = logit-transformed Wald-type confidence interval; IQR = interquartile range; NA = not applicable; PATH = Population Assessment of Tobacco and Health.
TABLE D-3 Tobacco Product Characteristics and Purchasing Behaviors Among Adult Current Established Traditional Cigar (Premium, Nonpremium), Cigarillo, Filtered Cigar and Cigarette Smokers, PATH Study Wave 4, 2016–2017
| Premium Cigars | Nonpremium Cigars | Cigarillos | Filtered Cigars | Cigarettesa | |
|---|---|---|---|---|---|
| % (95% CI) | % (95% CI) | % (95% CI) | % (95% CI) | % (95% CI) | |
| Has a regular brand | 38.4 (31.8–45.4) | 73.8 (65.3–80.9) | 74.8 (70.2–78.8) | 68.0 (59.6–75.4) | 92.5 (92.0–93.1) |
| Top 5 brands smokedb | Cohiba 20.3 | Black & Mild 23.6 | Black & Mild 51.0 | Black & Mild 20.7 | Marlboro 34.2 |
| (14.0–26.4) | (16.0–31.2) | (46.8–55.2) | (14.8–26.7) | (32.6–35.9) | |
| Arturo Fuente 17.9 | Acid 12.7 | Swisher Sweets 21.1 | Cheyenne 18.5 | Newport 15.9 | |
| (11.4–24.5) | (8.0–17.5) | (16.9–25.4) | (12.3–24.8) | (14.6–17.2) | |
| Romeo y Julieta 7.8 | Dutch Masters 11.8 | Backwoods 5.3 | Swisher Sweets 7.6 | Camel 10.8 | |
| (4.1–11.5) | (6.6–17.1) | (3.5–7.1) | (4.8–10.4) | (9.8–11.7) | |
| Montecristo 6.7 | Swisher Sweets 11.6 | White Owl 5.0 | 305’s 7.3 | Pall Mall 7.0 | |
| (4.0–9.3) | (7.7–15.5) | (3.0–7.0) | (2.8–11.8) | (6.1–7.8) | |
| Macanudo 5.2 | White Owl 5.9 | Dutch Masters 2.4 | Djarum 6.5 | American Spirit 3.6 | |
| (2.8–7.6) | (3.1–8.7) | (1.0–4.0) | (4.8–8.2) | (3.2–4.0) |
| Regular brand flavored or mentholatedb,c,d | 7.6 (5.5–10.5) | 50.3 (38.3–62.3) | 55.7 (50.4–60.9) | 48.4 (41.2–55.7) | 38.4 (36.8–40.0) |
| Usually buy in person Where buy tobacco producte | 76.4 (69.3–82.3) | 84.0 (74.9–90.3) | 92.5 (89.9–94.5) | 90.0 (85.5–93.1) | 94.9 (94.4–95.4) |
| Cigar bar | 35.4 (28.4–43.1) | * | * | * | NA |
| Convenience store/gas station | * | 64.5 (57.4–71.0) | 85.2 (81.8–88.1) | 73.7 (68.6–78.3) | 88.1 (87.2–88.9) |
| Smoke shop/tobacco specialty or outlet store | 44.3 (35.2–53.8) | * | 11.3 (9.0–14.2) | 22.1 (18.4–26.2) | 10.0 (9.0–11.2) |
| Somewhere else | * | * | * | * | 1.9 (1.3–2.6) |
| Usual purchase sizef | |||||
| Single | 71.5 (66.7–75.9) | 41.9 (35.1–49.0) | 52.1 (48.4–55.8) | 79.4 (74.4–83.7) | 2.5 (2.1–2.9) |
| Box or pack | 28.5 (24.1–33.3) | 58.1 (51.0–64.9) | 47.9 (44.2–51.6) | 20.6 (16.4–25.6) | 83.3 (82.0–84.5) |
| Carton | NA | NA | NA | NA | 14.2 (13.1–15.4) |
| Price per stickf,g (median, IQR) | $7.00 (4.00–10.00) | $1.07 (0.71–2.00) | $1.00 (0.66–1.08) | $0.11 (0.07–0.26) | $0.30 (0.25–0.35) |
a When a respondent reported smoking both manufactured cigarettes and RYO cigarettes (n = 753), for certain topics, they were asked separate questions about each product. For dual manufactured cigarette and RYO smokers, the responses to manufactured cigarette products are provided; otherwise, responses reflect the single cigarette type the respondent reported smoking.
b Among those with a regular brand; if no regular brand, refers to last brand purchased.
c Cigar and RYO smokers were asked whether their regular brand was flavored to taste like menthol, mint, clove, spice, candy, fruit, chocolate, alcohol, or sweets.
d Manufactured cigarette and RYO smokers were asked whether their regular brand was mentholated.
e Only asked of those who usually buy in person. Where buy tobacco product refers to where purchasing most of the time. “Convenience store/gas station” category also includes supermarket, grocery store, warehouse, or liquor store; “somewhere else” category also includes duty free shop, military commissary, bar/pub, restaurant, casino, friend, relative, swap meet/flea market, or store on an Indian reservation.
f For cigar smokers, restricted to usually buy in person; for cigarettes, asked of manufactured cigarette smokers, irrespective of buying in person or not.
g Among filtered cigar smokers, price per stick is restricted to those who reported purchasing either 20- or 12-count packs (44% of all filtered cigar smokers).
*The estimate has been suppressed because it is statistically unreliable. It is based on a (denominator) sample size of less than 50, or the relative standard error of the estimate (or its complement) is larger than 30 percent.
NOTE: CI = logit-transformed Wald-type confidence interval; IQR = interquartile range; NA = not applicable; PATH = Population Assessment of Tobacco and Health; RYO = roll-your-own.
TABLE D-4 Percent of Dual Cigar and Cigarette Smokers Among Adult Current Established Cigar Smokers and Adjusted Prevalence Ratios by Demographic and Cigar Smoking Characteristics, PATH Study Wave 4, 2016–2017
| Prevalencea (95% CI) (n = 1,688) | Adjusted PRa (95% CI) (n = 1,628) | |
|---|---|---|
| Smoke premium cigars | ||
| Yes | 23.2 (17.3–30.4) | 0.53 (0.41–0.69) |
| No | 59.8 (57.1–62.3) | Ref |
| Smoke nonpremium cigars | ||
| Yes | 46.5 (39.1–54.0) | 0.80 (0.67–0.95) |
| No | 51.5 (48.6–54.4) | Ref |
| Smoke cigarillos | ||
| Yes | 60.3 (57.2–63.2) | 1.37 (1.17–1.59) |
| No | 40.1 (36.2–44.0) | Ref |
| Smoke filtered cigars | ||
| Yes | 70.5 (64.9–75.6) | 1.58 (1.41–1.77) |
| No | 42.3 (39.3–45.4) | Ref |
| Use other tobacco productsb | ||
| Yes | 59.6 (53.5–65.4) | 1.14 (1.04–1.25) |
| No | 47.7 (45.0–50.4) | Ref |
| Sex | ||
| Male | 47.7 (45.2–50.2) | Ref |
| Female | 60.9 (56.9–64.8) | 1.05 (0.97–1.13) |
| Prevalencea (95% CI) (n = 1,688) | Adjusted PRa (95% CI) (n = 1,628) | |
|---|---|---|
| Age group (years) | ||
| 18–34 | 54.1 (50.1–58.2) | 1.01 (0.91–1.12) |
| 35+ | 48.1 (44.6–51.6) | Ref |
| Race/ethnicity | ||
| White, non-Hispanic | 50.0 (45.9–54.1) | Ref |
| Black/AA, non-Hispanic | 50.7 (45.4–56.1) | 0.85 (0.74–0.97) |
| Other/multi-race, or Hispanic | 53.7 (44.2–63.0) | 0.93 (0.80–1.09) |
| Education | ||
| GED, high school diploma, or less | 64.0 (60.1–67.7) | 1.38 (1.26–1.51) |
| Some college/associate degree or more | 38.2 (34.7–41.7) | Ref |
| Daily cigar smokingc | ||
| Yes | 50.9 (44.3–57.4) | 0.82 (0.72–0.92) |
| No | 50.6 (46.9–54.4) | Ref |
a There were n = 1,688 current established cigar smokers with information on current cigarette smoking status. The regression analysis included n = 1,628 participants (n = 774 cigar only; n = 854 dual cigar + cigarette) after observations missing information for covariate were excluded.
b Use of other tobacco products defined as having ever used one or more of the following tobacco products “fairly regularly” and now using that product every day or some days: ENDS, pipe tobacco, hookah, smokeless tobacco, or snus.
c Daily cigar refers to smoking at least one cigar type on a daily basis.
NOTE: AA = African American; CI = logit-transformed Wald-type confidence interval; PATH = Population Assessment of Tobacco and Health; PR = prevalence ratio.
TABLE D-5 Tobacco Dependence among Current Established Exclusivea Users of Four Cigar Types, Cigarette Smokers, Users of Smokeless Tobacco and Hookah, and the Overall Adult Population in U.S. PATH Adults
| Premium Cigars | Nonpremium Cigars | Cigarillos | Filtered Cigars | Cigarettes | Smokeless Tobacco | Hookah | ENDS | |
|---|---|---|---|---|---|---|---|---|
| Sample size, n | ||||||||
| Wave 1 | 165 | 70 | 289 | 79 | 8,223 | 945 | 451 | 415 |
| Wave 2 | 145 | 46 | 162 | 60 | 7,118 | 776 | 258 | 518 |
| Wave 3 | 160 | 37 | 152 | 56 | 6,828 | 712 | 217 | 620 |
| Wave 4 | 181 | 58 | 254 | 77 | 7,663 | 799 | 209 | 779 |
| Wave 5 | 177 | 45 | 262 | 72 | 6,387 | 644 | 160 | 1,639 |
| Mean TDS (95% CI) | ||||||||
| Wave 1 | 7.8 (6.4–9.1) | 13.1 (10.4–15.9) | 22.4 (20.1–24.8) | 35.2 (30.7–39.6) | 52.4 (51.8–53.0) | 45.4 (44.1–46.6) | 8.1 (7.4–8.8) | 25.7 (23.8–27.7) |
| Wave 2 | 7.1 (5.7–8.5) | 12.9 (10.2–15.6) | 21.2 (19.1–23.3) | 31.5 (27.6–35.3) | 49.6 (48.9–50.3) | 42.9 (41.8–43.9) | 4.9 (3.8–6.0) | 21.1 (19.3–22.8) |
| Wave 3 | 6.0 (4.3–7.7) | 16.9 (8.0–25.8) | 17.1 (13.8–20.4) | 36.3 (27.1–45.4) | 50.6 (50.1–51.0) | 44.1 (42.4–45.9) | 5.2 (4.1–6.3) | 32.9 (31.2–34.6) |
| Wave 4 | 6.6 (5.4–7.8) | 10.4 (7.3–13.4) | 20.3 (17.4–23.2) | 36.6 (29.2–43.9) | 51.1 (50.6–51.5) | 44.0 (41.9–46.0) | 6.8 (5.1–8.6) | 36.7 (34.6–38.8) |
| Wave 5 | 4.8 (4.0–5.6) | 16.7 (9.4–23.9) | 20.2 (17.9–22.4) | 34.7 (27.6, 41.8) | 51.4 (50.7–52.2) | 45.7 (44.4–47.1) | 7.6 (6.3–8.9) | 37.1 (35.6–38.6) |
| Median TDS (IQR) | ||||||||
| Wave 1 | 3.1 (0.0–9.4) | 3.1 (0.0–15.6) | 12.5 (3.1–35.5) | 31.3 (9.4–50.2) | 53.1 (31.3–75.0) | 43.8 (21.9–68.8) | 3.1 (0.0–9.4) | 18.8 (3.1–43.3) |
| Wave 2 | 3.1 (0.0–9.4) | 6.3 (0.0–18.8) | 12.5 (3.1–34.4) | 27.9 (9.4–46.9) | 50.0 (28.1–71.9) | 40.6 (18.8–62.5) | 3.1 (0.0–6.3) | 12.5 (0.0–35.2) |
| Wave 3 | 0.0 (0.0–6.3) | 3.1 (0.0–16.8) | 9.0 (0.0–28.1) | 33.6 (4.8–56.3) | 53.1 (28.1–73.3) | 46.9 (18.8–65.6) | 0.0 (0.0–6.3) | 31.3 (6.3–53.1) |
| Wave 4 | 0.0 (0.0–6.3) | 0.0 (0.0–9.4) | 12.5 (0.0–31.3) | 34.7 (9.4–56.8) | 53.1 (28.1–75.0) | 43.8 (18.8–65.6) | 0.0 (0.0–6.3) | 34.4 (9.4–56.3) |
| Wave 5 | 0.0 (0.0–6.3) | 3.1 (0.0–24.8) | 12.5 (3.1–31.3) | 21.5 (6.3–64.2) | 53.1 (31.3–75.0) | 46.9 (18.8–68.8) | 0.0 (0.0–9.4) | 34.3 (12.5–56.3) |
| Premium Cigars | Nonpremium Cigars | Cigarillos | Filtered Cigars | Cigarettes | Smokeless Tobacco | Hookah | ENDS | |
|---|---|---|---|---|---|---|---|---|
| Percentage of report 1+ symptoms (95% CI): TDS >0 | ||||||||
| Wave 1 | 59.6 (52.9–65.8) | 63.8 (51.4–74.6) | 76.5 (71.9–80.5) | 92.5 (89.5–94.7) | 96.1 (95.6–96.4) | 95.6 (94.3–96.6) | 63.0 (60.8–65.1) | 79.4 (75.7–82.7) |
| Wave 2 | 59.7 (50.6–68.2) | 67.8 (54.4–78.8) | 78.6 (71.3–84.4) | 87.3 (80.4–92.1) | 95.2 (94.6–95.6) | 94.0 (90.8–96.1) | 50.1 (45.1–55.2) | 71.8 (67.5–75.7) |
| Wave 3 | 49.1 (39.1–59.3) | 64.0 (53.6–73.3) | 70.8 (62.6–77.9) | 81.7 (72.5–88.3) | 94.5 (93.6–95.2) | 91.1 (88.9–93.0) | 44.3 (36.4–52.4) | 82.8 (80.2–85.1) |
| Wave 4 | 46.9 (41.5–52.4) | 46.6 (34.8–58.7) | 73.0 (69.0–76.6) | 85.6 (76.0–91.8) | 95.2 (94.7–95.6) | 94.7 (93.4–95.8) | 47.5 (43.4–47.5) | 90.0 (88.3–91.5) |
| Wave 5 | 43.3 (36.1–50.9) | 54.2 (41.7–66.1) | 75.5 (67.2–82.3) | 87.2 (77.0–93.2) | 94.7 (93.9–95.4) | 95.8 (93.7–97.2) | 48.4 (42.8–54.1) | 89.1 (87.7–90.4) |
| Sample size and difference in mean TDS between nondaily and daily users (95% CI) | ||||||||
| Wave 1 | 151/14 26.9 (21.3–32.6) | 53/17 20.1 (10.0–30.2) | 209/80 21.9 (15.0–28.9) | 43/36 32.3 (24.2–40.5) | 1607/6616 37.6 (36.6–38.5) | 248/691 27.2 (24.7–29.7) | 434/17 19.1 (14.5–23.8) | 147/268 13.1 (10.7–15.5) |
| Wave 2 | 135/10 15.6 (6.1–25.1) | 39/7 24.2 (15.0–33.5) | 119/43 20.5 (15.2–25.7) | 35/25 22.3 (15.7–28.9) | 1425/5693 35.8 (34.9–36.7) | 201/575 30.2 (27.3–33.1) | 246/12 27.0 (13.9–40.0) | 161/357 14.1 (11.7–16.4) |
| Wave 3 | 153/7 25.9 (8.2–43.7) | 29/8 24.3 (5.0–43.7) | 105/47 17.5 (9.7–25.2) | 30/26 25.7 (17.0–34.5) | 1266/5562 36.7 (35.6–37.8) | 170/542 34.2 (32.2–36.3) | 211/6 16.2 (1.1–31.2) | 221/399 30.5 (28.1–33.0) |
| Wave 4 | 172/9 13.5 (–4.7–31.6) | 40/18 20.8 (8.0–33.6) | 183/71 24.1 (18.9–29.3) | 40/37 17.1. (4.4–29.9) | 1561/6102 36.6 (35.7–37.5) | 205/594 27.6 (23.8–31.4) | 205/4 40.4 (34.3–46.5) | 260/519 29.4 (26.0–32.8) |
| Wave 5 | 166/11 9.4 (–2.8–21.6) | 38/7 13.8 (0.8–26.8) | 177/85 19.9 (13.1–26.7) | 36/36 42.9 (34.9–50.9) | 1304/ 5,083 36.6 (35.7–37.5) | 163/481 28.3 (25.3–31.4) | 151/9 28.8 (19.5–38.0) | 633/1006 29.4 (27.4–31.4) |
| Premium Cigars | Nonpremium Cigars | Cigarillos | Filtered Cigars | Cigarettes | Smokeless Tobacco | Hookah | ENDS | |
|---|---|---|---|---|---|---|---|---|
| Sample size and difference in mean TDS between users of <6 days and 6+ days used in the past 30 daysb (95% CI) | ||||||||
| Wave 1 | 123/40 14.6 (10.2–19.0) | 34/34 13.1 (5.9–20.2) | 101/178 20.0 (15.8–24.2) | 22/54 20.9 (11.1–30.7) | 526/7633 43.4 (42.0–44.8) | 70/832 31.1 (26.6–35.6) | 163/84 9.1 (5.3–13.0) | 66/345 15.6 (12.0–19.2) |
| Wave 2 | 117/25 9.0 (2.6–15.3) | 28/16 16.1 (7.9–24.4) | 72/85 20.1 (17.7–22.6) | 23/35 13.9 (4.0–23.8) | 436/6619 41.4 (39.7–43.1) | 65/686 35.5 (31.0–40.1) | 75/42 2.9 (0.7–5.1) | 84/429 13.3 (9.5–17.1) |
| Wave 3 | 128/27 11.4 (6.5–16.4) | 20/16 25.4 (3.8–47.0) | 70/80 14.3 (8.8–19.8) | 20/36 18.9 (9.5–28.2) | 439/6344 38.8 (36.3–41.2) | 64/626 35.1 (30.6–39.7) | 64/19 12.0 (8.8–15.3) | 102/510 26.5 (22.4–30.7) |
| Wave 4 | 144/33 12.6 (4.0–21.2) | 29/28 15.7 (7.0–24.3) | 97/151 17.1 (13.0–21.1) | 27/50 13.9 (–2.6–30.5) | 563/7044 39.2 (37.6–40.7) | 65/701 29.6 (23.9–35.4) | 62/33 13.8 (1.9–25.7) | 120/652 30.2 (26.3–34.2) |
| Wave 5 | 145/30 5.9 (1.2–10.6) | 30/15 10.9 (0.3–21.5) | 95/163 17.6 (13.4–21.8) | 21/51 30.9 (22.6–39.1) | 499/5852 42.3 (40.5–44.1) | 53/564 32.5 (29.1–35.9) | 34/31 10.2 (3.4–17.0) | 267/1362 29.4 (27.6–31.3) |
| Linear association of the number of days used in the past 30 daysb with TDS (95% CI) | ||||||||
| Wave 1 | 0.9 (0.7–1.1) | 0.7 (0.4–1.0) | 0.9 (0.7–1.1) | 1.1 (0.7–1.4) | 1.8 (1.8–1.9) | 1.4 (1.2–1.5) | 0.0 (0.0–0.0) | 0.6 (0.5–0.7) |
| Wave 2 | 0.5 (0.3–0.8) | 0.9 (0.6–1.2) | 0.9 (0.8–1.0) | 0.7 (0.4–1.1) | 1.8 (1.7–1.8) | 1.4 (1.3–1.6) | 0.0 (0.0–0.2) | 0.6 (0.5–0.7) |
| Wave 3 | 0.6 (0.2–0.9) | 1.5 (0.8–2.3) | 0.6 (0.4–0.9) | 0.9 (0.5–1.3) | 1.7 (1.7–1.8) | 1.5 (1.4–1.7) | 0.5 (0.2–0.9) | 1.3 (1.1–1.4) |
| Wave 4 | 0.6 (0.0–1.1) | 0.7 (0.3–1.1) | 0.9 (0.7–1.0) | 0.6 (0.2–1.1) | 1.7 (1.6–1.7) | 1.3 (1.1–1.6) | 0.2 (–0.1–0.5) | 1.3 (1.2–1.4) |
| Wave 5 | 0.3 (0.0–0.7) | 0.8 (0.4–1.2) | 0.7 (0.5–0.9) | 1.5 (1.2–1.8) | 1.7 (1.7–1.8) | 1.5 (1.3–1.7) | 0.6 (0.4–0.7) | 1.3 (1.2–1.4) |
a Exclusive means that they use only that product currently or in past 30 days but no other tobacco product. For cigar users, exclusive also eliminates respondents who report using a second cigar product type.
b For hookah users, the frequency information was reported by the number of “times” instead of “days” used in the past 30 days in PATH data.
NOTE: CI = logit-transformed Wald-type confidence interval; ENDS = electronic nicotine delivery systems; IQR = interquartile range; PATH = Population Assessment of Tobacco and Health; TDS = tobacco dependence score.
NOTE: PATH = Population Assessment of Tobacco and Health.
NOTE: PATH = Population Assessment of Tobacco and Health.
NOTE: PATH = Population Assessment of Tobacco and Health.
Supplemental Table A Assignment of Adult Current Established Traditional Cigar Smokers as Either Premium or Nonpremium Based on Usual Cigar Brand, PATH Study, Waves 1–5
| Since regulatory definitions of premium cigars do not exist, information about the brand’s tobacco blends, components (e.g., long filler, whole leaf wrapper), and manufacturing process (e.g., handmade), obtained through online searches, was used to distinguish premium from nonpremium brands. | |
| Assignment | Brand Name Reported |
| Premium cigar brands | 5 Star, 5 Vegas, Alec Bradley, Arturo Fuente, Ashton, Ben-Bay, Bolivar, Brickhouse, Buccanero, CAO, Casablanca, Chubb, Churchill, Cohiba, CubaLibre, Cuban, Davidoff, Don Tomas, Drew Estates, Durango, El Pita, Elverso, Excalibur, Flor de Oliva, Gurkha, H. Uppmanns, Hoya de Monterrey, Indio, Joya de Nicaragua, Kristoff, La Corona, La Gloria Cubana, Los Blancos, Macanudo, Makers Mark, Man of War, Montecristo, My Father, Nub, Oliva, Padron, Partagas, Perdomo, Professor Sila, Punch, Robert Burns Blackwatch, Rocky Patel, Romeo y Julieta, San Cristo, Tatuaje, Torano, Victor Sinclair, Zeno |
| Nonpremium cigar brands | Acid, Al Capone, Antonio y Cleopatra, Backwoods, Black and Mild, Blackstone, BLK, Djarum, Dutch Masters, El Producto, Entourage, Game, Garcia y Vega, Hav-A-Tampa, Isla Del Sol, Java, Kahula, King Edward, Marsh Wheeling, Miami Suites, Middletons, Munnimaker, Muriels, Optimo, Parodi, Phillies, Spliterillo, Supre Sweets, Supreme Blend Peach Cigars, Supreme Menthol, Swisher Sweets, Tampa Nugget, Tampa Sweet, White Owl |
|
|
NOTE: PATH = Population Assessment of Tobacco and Health.
