Flavor Types Used by Youth and Adult Tobacco Users in Wave 2 of the Population Assessment of Tobacco and Health (PATH) Study 2014-15

Abstract

Most youth and young adult (YA) tobacco users use flavored products; however, little is known about specific flavors used. We report flavor types among U.S. tobacco users from the Population Assessment of Tobacco and Health Study, Wave 2, 2014–2015.

At Wave 2, we examined (1) flavor use and type at past 30-day use; (2) new flavored tobacco product use and type; (3) product-specific flavor patterns across youth (ages 12–17) (n=920), YA (18–24) (n=3,726), and adult (25+) (n=10,346) past 30-day and new tobacco users; and (4) concordance between self- and expert-coded brand flavor type among all adults (18+).

Prevalence of flavored tobacco product use was highest among youth, followed by YA and adult 25+ any tobacco users. Within each age group, flavored use was greatest among hookah, e-cigarette, and snus users. Overall, menthol/mint, fruit, and candy/sweet were the most prevalent flavor types at first and past 30-day use across age groups. For past 30-day use, all flavor types except menthol/mint exhibited an inverse age gradient, with more prevalent use among youth and YAs, followed by adults 25+. Prevalence of menthol/mint use was high (over 50% youth, YAs; 76% adults 25+) and exhibited a positive age gradient overall, though the reverse for cigarettes. Brand-categorized and self-reported flavor use measures among adults 18+ were moderately to substantially concordant across most products.

Common flavors like menthol/mint, fruit, and candy/sweet enhance appeal to young tobacco users. Information on flavor types used by product and age can inform tobacco flavor regulations to addess flavor appeal especially among youth.

Introduction

In 2009, U.S. Federal law banned characterizing flavors except menthol in cigarettes, but not in non-cigarette products.¹ Some U.S. localities have enacted sales restrictions on a broader range of flavored tobacco products (FTPs).² Nonetheless, U.S. national data show that FTP use is common, with a strong inverse age gradient and higher use among young tobacco users.^3–12 Almost 80% of current youth tobacco users used an FTP, compared with half of adult users.¹³ Additionally, over 80% of youth^{3, 13} and 55% of adults who ever used tobacco used a FTP at first use, and such use was associated with current use.¹³ The prevalence of current FTP use varies by product, with highest rates found for hookah, e-cigarettes, smokeless tobacco, and cigars.^{3, 4, 6, 8, 9, 12, 14, 15}

Tobacco industry documents reveal that menthol cigarettes and tobacco products with sweet or spicy flavors were developed to mask the harsh taste of tobacco¹⁶ and appeal to new initiates and younger smokers.^17–21 Understanding use of specific flavors is important as FTPs continue to be marketed to young populations in the retail environment, online, and through social media.^22–25 Sweet and fruit flavors in cigarettes, cigars, e-cigarettes, and hookah appeal across ages to new users.^{19–21, 26–29} National Adult Tobacco Survey data (2013–2014) showed differing flavors used across products, with menthol/mint flavors popular among smokeless tobacco users, fruit flavors among hookah and cigar users, and menthol/mint and fruit flavors among adult 18+ e-cigarette users.⁸ However, no assessment has compared flavor types used across a broad range of tobacco products among youth, YAs, and adults 25+, in a nationally representative study.

FTP sales grew for cigarillos, chewing tobacco, menthol cigarettes, little cigars, moist snuff, and snus between 2011 and 2015,³⁰ and for products with ambiguous descriptors, like “Jazz” cigars,³¹ “Purple K” cigarillos,³² and “Unicorn Milk” e-cigarettes.³³ Little is known about how users characterize the flavor type of tobacco products with ambiguous descriptors, and no study has examined concordance between self-reported FTP use and brand flavor use as characterized by an expert panel based on respondent brand.

This study assessed flavor types used by youth (ages 12–17), YAs (ages 18–24), and adults ages 25+, and differences by product and age. Specific objectives were to examine (1) flavor use and type at past 30-day use at Wave 2; (2) new FTP use between Waves 1 and 2; (3) product-specific flavor patterns across youth (n = 920), YA (n = 3,726) and adult 25+ (n = 10,346) past 30-day and new tobacco users; and (4) concordance between self- and expert-coded brand flavor type among all adults (ages 18+).

Methods

The National Institutes of Health, National Institute on Drug Abuse with the Food and Drug Administration’s Center for Tobacco Products, conducts the PATH Study under a contract with Westat. The PATH Study uses audio computer-assisted self-interviews in English and Spanish. Wave 1 data collection occurred between September 12, 2013, and December 14, 2014 and Wave 2 between October 23, 2014, and October 30, 2015. Wave 2 respondents were interviewed approximately 1 year after their Wave 1 interview. Recruitment employed a stratified address-based, area-probability sampling design, with oversampling of adult tobacco users, YAs, and African-American adults.

Population and replicate weights were adjusted for the complex study design and nonresponse at Waves 1 and 2. With a probability sample, weights allow for the computation of representative estimates among the non-institutionalized, civilian U.S. population ages 12+. The PATH Study was approved by Westat’s Institutional Review Board. Design and methods are published elsewhere.³⁴ Survey procedures, questionnaires, sampling, weighting, and data access are available at https://doi.org/10.3886/Series606. All participants ages 18+ provided informed consent. Youth participants provided assent; their parent/legal guardian provided consent.

The current study analyzed cross-sectional data from Wave 2 Youth and Adult Interviews. At Wave 2, 28,362 adults and 12,172 youth completed interviews; the weighted response rates were 83.2% (Adult Interview) and 87.3% (Youth Interview).Wave 2 included 2,091 ‘aged up’ youth at Wave 1 who had reached age 12 by Wave 2 and were interviewed for the first time, and 1,915 youth at Wave 1 who had reached age 18 by Wave 2 and responded to the Adult Interview (see PATH Study User Guide at https://doi.org/10.3886/Series606).

Measures

Table S1 defines all constructs measured in this study.

Tobacco Product Use

The PATH Study assessed youth and adult Wave 2 past 30-day tobacco use for cigarettes, e-cigarettes (disposable, non-refillable or refillable cartridge, or tank), traditional (large) cigars, cigarillos, filtered (little) cigars, any cigar,^a hookah, smokeless tobacco (loose snus, snuff, dip, chew, and spit, excluding snus pouches), snus pouches (hereafter termed ‘snus’), pipe, and dissolvable tobacco (only adults 18+ were asked about use of pipe and dissolvable tobacco). At Wave 2, participants were asked about first tobacco product use and were defined as new product users if they had tried a new tobacco product between Waves 1 and 2. This included never-tobacco users at Wave 1 who used any tobacco product at Wave 2, and ever-tobacco users at Wave 1 who used a new product between Waves 1 and 2.

Brand/Sub-Brand

For each tobacco product, adult established users^b (current users and recent former users, i.e., users at Wave 1 not currently using at Wave 2) were asked to identify their usual or last brand (e.g., Newport) by selecting a brand from a list with pictures or by writing in a response. Respondents selecting a brand could select a specific sub-brand from a series of pictures (e.g., Newport Full-Flavor Menthol Kings) or write-in a response.

Flavored Tobacco Product (FTP) Use and Flavor Type

For each tobacco product, we assessed FTP use and flavor type with respect to (1) first use of each product (among new product users), (2) past 30-day product use, and (3) product brand (among adult established users). Respondents using new tobacco products between Waves 1 and 2 reported whether first use of the product was “flavored to taste like menthol, mint, clove, spice, candy, fruit, chocolate, alcohol (such as wine or cognac), or other sweets” (yes/no), (first FTP use). If yes, respondents answered: “Which flavor did you first start using/smoking?” (menthol/mint, clove or spice, fruit, chocolate, alcoholic drink [e.g., wine, cognac, margarita/other cocktails], candy/other sweets, or some other flavor) (first FTP type). Given that flavors could potentially fall into multiple categories (e.g., “strawberry daiquiri” could be considered as both fruit and alcoholic drink), respondents were asked to select all applicable flavors. First FTP users of any tobacco product indicated at least one new product was flavored. We classified new product users of specific flavor types (e.g., first menthol user) if any first FTP used was that flavor. If Wave 2 respondents used more than one new product since Wave 1 (n = 1,812), we could not determine their first new product; however, these responses are included in analyses for any product and each new product used.

Youth and adult past 30-day tobacco users at Wave 2 reported whether each tobacco product used in the past 30 days was flavored (past 30-day flavor use) and, if so, “Which flavors have you smoked/used in the past 30 days?” (past 30-day flavor type), with the same flavor options as above. Past 30-day flavor users used at least one FTP in the past 30 days. We classified past 30-day users of specific flavor types (e.g., past 30-day fruit user) if any tobacco product used in the past 30 days was that flavor.

Respondents answered whether their brand was flavored (brand flavor use) and its flavor type (brand flavor type) using similar questions as described above (adult current and recent former established users only).

Expert-coded Flavor Type and Primary Flavor

For the concordance analysis, an expert panel (SWR, AG, OW, CD, MBT, ACV) coded respondent sub-brands into expert-coded flavor types of menthol/mint, clove/spice, fruit, chocolate, alcohol, candy/sweet, and other flavor or as non-flavored (i.e., tobacco) or unknown flavor (ambiguous descriptors). Sub-brands were coded based on flavor descriptors in the names, on the product packaging/pictures, or on descriptors on manufacturers’ websites. The expert panel independently coded descriptors, initially agreeing on the coding for 83% of descriptors and reaching consensus on all final coding after discussion. Expert-coded flavor type was coded as “check all that apply,” resulting in one or two flavor types. Expert-coded primary flavor was coded for sub-brands with more than one flavor type by assigning flavors in the following order: menthol/mint, alcohol, fruit, candy/sweet, chocolate, clove/spice, and/or other flavor. This ordering was based on expert panel consensus (e.g., “strawberry daiquiri” was given a primary flavor of alcohol, not fruit).

Statistical Analyses

Analyses used SVY procedures in Stata/SE version 14.1 to account for weighting. Weighted population-based prevalence estimates are presented for flavor use and type, with 95% confidence intervals using the balanced repeated replication method with Fay’s adjustment. For each tobacco product type, differences were tested by age group, at first use (Table 1) and past 30-day use (Table 2), with chi-square tests for flavor type prevalence among FTP users. P-values were adjusted using the Benjamini-Hochberg method for comparisons across multiple age groups; analyses were conducted in R version 3.3.1.^c For estimates of FTP use consistent with prior Wave 1 analyses,^{3, 13} missing values were excluded for use of an FTP (refused/don’t know) from the analytic sample; however, missing values for flavor type were retained in the denominator for flavor use (see Table 1 and 2 footnotes for details of missingness). “No” or “I don’t know” responses were categorized as non-flavored use. Product-specific analyses (Tables S2 and S3) examined percentages of self-reported flavor type among all new or past 30-day users, instead of among FTP users.

Table 1.

Use of a flavored tobacco product and flavor of first product used among new flavored product users by age group^† and tobacco product^‡ in Wave 2

		First use among new users from Wave 1 to Wave 2 and First Flavor Type among New Flavored Users§
		Youth (12–17)			Young Adults (18–24)			Adults (25+)
		Frequency	Weighted Percentage	95% CI	Frequency	Weighted Percentage	95% CI	Frequency	Weighted Percentage	95% CI	Hypothesis Test
Any Tobacco Product	New Product Users Wave 1 to Wave 2a	1362	11.4%b	(10.8, 12.1)	2456	26.9%c	(25.6, 28.3)	2745	8.1%d	(7.7, 8.5)
	Non-flavored	362	25.7%	(23.4, 28.2)	1025	40.4%	(37.9, 42.9)	1431	54.8%	(52.2, 57.5)
	Any flavor usee	1000	74.3%	(71.8, 76.6)	1431	59.5%	(57.0, 62.0)	1314	45.1%	(42.5, 47.8)	g, h, i*
	Menthol or Mintf	302	30.4%	(27.4, 33.7)	571	39.6%	(36.8, 42.4)	593	44.6%	(41.1, 48.2)	g, h, i*
	Clove or Spicef	21	1.9%	(1.3, 2.9)	71	4.3%	(3.4, 5.5)	79	5.8%	(4.3, 7.8)	g,h*
	Fruitf	564	56.7%	(53.6, 59.7)	838	60.4%	(57.6, 63.2)	520	39.2%	(36.0, 42.5)	h,i*
	Chocolatef	72	6.9%	(5.3, 8.9)	127	8.6%	(7.1, 10.4)	87	6.3%	(4.8, 8.1)
	Alcoholic Drinkf	43	4.3%	(3.1, 5.9)	139	9.9%	(8.2, 11.8)	100	7.2%	(5.6, 9.3)	g, h*
	Candy/Other Sweetsf	276	27.5%	(25.1, 30.1)	442	30.9%	(27.6, 34.4)	236	18.8%	(16.3, 21.5)	h, i*
	Other Flavorf	134	13.4%	(11.4, 15.6)	194	13.6%	(11.6, 16.0)	151	10.8%	(8.7, 13.3)
Cigarettes	New Product Users Wave 1 to Wave 2	335	2.8%	(2.5, 3.1)	336	4.0%	(3.5, 4.6)	79	0.4%	(0.3, 0.5)
	Non-flavored	203	59.7%	(53.3, 65.7)	216	63.1%	(57.7, 68.3)	46	62.4%	(47.3, 75.5)
	Any flavor use	132	40.4%	(34.3, 46.7)	120	36.9%	(31.8, 42.3)	33	37.6%	(24.5, 52.7)
	Menthol or Mint	85	65.1%	(55.6, 73.6)	105	88.3%	(80.2, 93.4)	-	-	-	g
	Clove or Spice	-	-	-	-	-	-	-	-	-
	Fruit	24	17.0%	(10.9, 25.5)	-	-	-	-	-	-
	Chocolate	-	-	-	-	-	-	-	-	-
	Alcoholic Drink	-	-	-	-	-	-	-	-	-
	Candy/Other Sweets	-	-	-	-	-	-	-	-	-
	Other Flavor	18	14.0%	(8.5, 22.1)	-	-	-	-	-	-
E-cigarettes	New Product Users from Wave 1 to Wave 2	809	6.8%	(6.2, 7.5)	905	10.5%	(9.7, 11.4)	1310	3.8%	(3.5, 4.1)
	Non-flavored	155	18.4%	(15.7, 21.5)	243	26.0%	(23.2, 29.1)	604	45.7%	(42.0, 49.4)
	Any flavor use	654	81.6%	(78.5, 84.3)	662	74.0%	(70.9, 76.8)	706	54.3%	(50.6, 58.0)	g,h, i*
	Menthol or Mint	91	13.3%	(10.5, 16.7)	167	25.0%	(21.3, 29.1)	295	39.5%	(34.9, 44.3)	g,h,i*
	Clove or Spice	-	-	-	-	-	-	23	3.2%	(2.0, 5.2)
	Fruit	384	59.1%	(55.2, 62.8)	417	65.2%	(61.2, 69.0)	286	41.3%	(36.9, 45.9)	g,h,i*
	Chocolate	40	5.9%	(4.1, 8.5)	49	6.8%	(5.0, 9.2)	34	4.7%	(3.1, 6.9)
	Alcoholic Drink	15	2.0%	(1.2, 3.4)	33	5.2%	(3.5, 7.5)	28	3.5%	(2.2, 5.4)	g*
	Candy/Other Sweets	192	29.2%	(26.0, 32.7)	249	38.7%	(34.2, 43.4)	148	21.4%	(17.7, 25.7)	g,h,i*
	Other Flavor	71	10.4%	(8.3, 13.0)	72	11.2%	(9.0, 14.0)	79	10.8%	(8.2, 14.1)
Traditional Cigars	New Product Users from Wave 1 to Wave 2	102	0.9%	(0.7, 1.1)	723	7.8%	(7.1, 8.5)	661	2.0%	(1.8, 2.2)
	Non-flavored	69	69.8%	(58.2, 79.4)	531	73.6%	(69.6, 77.3)	484	76.6%	(71.7, 80.9)
	Any flavor use	33	30.2%	(20.6, 41.8)	192	26.4%	(22.8, 30.4)	177	23.4%	(19.1, 28.3)
	Menthol or Mint	-	-	-	45	20.3%	(14.8, 27.2)	55	40.1%	(29.7, 51.5)	i*
	Clove or Spice	-	-	-	17	7.7%	(4.8, 12.1)	-	-	-
	Fruit	-	-	-	81	45.5%	(37.8, 53.4)	55	26.0%	(19.6, 33.5)	i*
	Chocolate	-	-	-	31	15.4%	(10.4, 22.4)	18	9.4%	(5.8, 14.8)
	Alcoholic Drink	-	-	-	30	15.7%	(10.2, 23.3)	21	10.7%	(6.7, 16.6)
	Candy/Other Sweets	-	-	-	34	16.3%	(11.7, 22.2)	31	16.6%	(11.0, 24.2)
	Other Flavor	-	-	-	32	17.6%	(11.8, 25.3)	24	13.9%	(8.1, 22.8)
Cigarillos	New Product Users from Wave 1 to Wave 2	151	1.3%	(1.1, 1.6)	343	3.6%	(3.1, 4.1)	296	0.8%	(0.7, 1.0)
	Non-flavored	76	49.2%	(41.8, 56.7)	178	51.1%	(45.2, 57.1)	172	58.6%	(50.4, 66.2)
	Any flavor use	75	50.8%	(43.3, 58.2)	165	48.9%	(42.9, 54.8)	124	41.5%	(33.8, 49.6)
	Menthol or Mint	-	-	-	27	14.0%	(9.8, 19.8)	36	35.4%	(24.4, 48.2)	i*
	Clove or Spice	-	-	-	-	-	-	-	-	-
	Fruit	38	50.2%	(38.9, 61.5)	100	64.8%	(56.1, 72.6)	41	28.7%	(19.1, 40.8)	h,i*
	Chocolate	-	-	-	14	7.9%	(4.7, 13.0)	-	-	-
	Alcoholic Drink	-	-	-	31	20.1%	(13.0, 29.6)	21	21.6%	(12.3, 35.3)
	Candy/Other Sweets	13	17.3%	(9.8, 28.5)	37	20.6%	(15.0, 27.7)	21	19.3%	(11.5, 30.6)
	Other Flavor	-	-	-	27	15.4%	(10.3, 22.5)	20	14.9%	(8.0, 26.0)
Filtered Cigars	New Product Users from Wave 1 to Wave 2	73	0.6%	(0.5, 0.8)	362	3.5%	(3.2, 3.8)	417	1.0%	(0.9, 1.2)
	Non-flavored	36	49.0%	(36.7, 61.4)	242	68.7%	(63.4, 73.6)	241	60.3%	(54.9, 65.3)
	Any flavor use	37	51.0%	(38.6, 63.3)	120	31.3%	(26.4, 36.6)	176	39.8%	(34.7, 45.1)	g
	Menthol or Mint	-	-	-	39	28.8%	(20.4, 38.9)	88	48.1%	(39.7, 56.6)	i*
	Clove or Spice	-	-	-	12	9.6%	(5.3, 16.8)	21	12.0%	(7.2, 19.4)
	Fruit	-	-	-	54	45.8%	(37.3, 54.5)	47	25.0%	(18.9, 32.2)	i*
	Chocolate	-	-	-	-	-	-	-	-	-
	Alcoholic Drink	-	-	-	15	12.8%	(7.5, 21.1)	18	12.2%	(7.4, 19.4)
	Candy/Other Sweets	-	-	-	18	16.6%	(9.3, 28.0)	26	17.2%	(10.8, 26.3)
	Other Flavor	-	-	-	-	-	-	-	-	-
Any Cigars	New Product Users from Wave 1 to Wave 2	252	2.2%	(1.9, 2.5)	1150	12.0%	(11.2, 12.8)	1121	3.2%	(2.9, 3.4)
	Non-flavored	125	49.6%	(42.9, 56.3)	739	64.8%	(61.6, 67.9)	704	66.9%	(62.8, 70.7)
	Any flavor use	127	50.4%	(43.7, 57.1)	411	35.2%	(32.2, 38.4)	417	33.2%	(29.3, 37.2)	g,h*
	Menthol or Mint	22	15.9%	(10.2, 24.1)	95	20.2%	(16.4, 24.6)	165	42.5%	(36.5, 48.7)	h,i*
	Clove or Spice	-	-	-	32	7.3%	(5.2, 10.2)	36	7.9%	(5.5, 11.4)
	Fruit	64	51.0%	(42.7, 59.2)	209	53.4%	(48.6, 58.2)	133	29.0%	(23.6, 35.0)	h,i*
	Chocolate	-	-	-	53	12.7%	(9.7, 16.6)	36	9.1%	(5.9, 13.8)
	Alcoholic Drink	-	-	-	73	18.1%	(14.1, 22.9)	57	15.4%	(10.9, 21.4)
	Candy/Other Sweets	17	13.8%	(9.0, 20.5)	84	19.6%	(15.8, 24.0)	67	18.5%	(13.8, 24.3)
	Other Flavor	22	17.5%	(12.1, 24.6)	64	15.3%	(11.8, 19.5)	50	11.0%	(7.4, 16.0)
Pipe	New Product Users from Wave 1 to Wave 2	N/Aj	N/A	N/A	138	1.4%	(1.2, 1.7)	93	0.3%	(0.2, 0.4)
	Non-flavored	N/A	N/A	N/A	106	77.6%	(67.3, 85.4)	63	76.6%	(66.6, 84.3)
	Any flavor use	N/A	N/A	N/A	32	22.4%	(14.6, 32.8)	30	23.4%	(15.7, 33.4)
	Menthol or Mint	N/A	N/A	N/A	-	-	-	-	-	-
	Clove or Spice	N/A	N/A	N/A	-	-	-	-	-	-
	Fruit	N/A	N/A	N/A	-	-	-	-	-	-
	Chocolate	N/A	N/A	N/A	-	-	-	-	-	-
	Alcoholic Drink	N/A	N/A	N/A	-	-	-	-	-	-
	Candy/Other Sweets	N/A	N/A	N/A	-	-	-	-	-	-
	Other Flavor	N/A	N/A	N/A	-	-	-	-	-	-
Hookah	New Product Users from Wave 1 to Wave 2	291	2.4%	(2.2, 2.8)	625	7.1%	(6.5, 7.8)	257	0.8%	(0.7, 0.9)
	Non-flavored	67	22.8%	(17.8, 28.8)	235	33.3%	(29.2, 37.6)	116	49.7%	(42.6, 56.8)
	Any flavor use	224	77.2%	(71.2, 82.2)	390	66.7%	(62.4, 70.8)	141	50.3%	(43.2, 57.4)	g,h,i*
	Menthol or Mint	49	21.1%	(16.3, 27.0)	108	29.5%	(23.9, 35.7)	29	18.8%	(12.2, 27.9)
	Clove or Spice	-	-	-	20	4.1%	(2.3, 7.2)	-	-	-
	Fruit	153	68.5%	(62.3, 74.1)	301	78.6%	(73.4, 83.0)	97	64.8%	(53.6, 74.6)	g,i*
	Chocolate	20	9.0%	(6.1, 13.1)	20	5.4%	(3.1, 9.5)	15	8.8%	(5.0, 15.1)
	Alcoholic Drink	-	-	-	31	8.2%	(5.1, 12.8)	14	7.5%	(4.2, 13.0)
	Candy/Other Sweets	67	31.0%	(25.1, 37.6)	136	31.0%	(25.9, 36.6)	26	16.0%	(10.1, 24.4)	h,i*
	Other Flavor	20	8.2%	(5.2, 12.7)	53	14.3%	(10.0, 19.9)	16	11.4%	(6.6, 19.1)
Smokeless Tobacco	New Product Users from Wave 1 to Wave 2	138	1.3%	(1.0, 1.5)	134	1.4%	(1.2, 1.7)	90	0.3%	(0.2, 0.3)
	Non-flavored	61	40.7%	(33.3, 48.6)	70	53.8%	(44.5, 62.9)	59	71.9%	(59.4, 81.7)
	Any flavor use	77	59.3%	(51.4, 66.7)	64	46.2%	(37.1, 55.5)	31	28.1%	(18.3, 40.6)	g,h,i*
	Menthol or Mint	63	79.5%	(69.5, 86.8)	57	90.0%	(77.6, 95.9)	-	-	-
	Clove or Spice	-	-	-	-	-	-	-	-	-
	Fruit	12	15.9%	(9.6, 25.2)	-	-	-	-	-	-
	Chocolate	-	-	-	-	-	-	-	-	-
	Alcoholic Drink	-	-	-	-	-	-	-	-	-
	Candy/Other Sweets	-	-	-	-	-	-	-	-	-
	Other Flavor	-	-	-	-	-	-	-	-	-
Snus	New Product Users from Wave 1 to Wave 2	69	0.6%	(0.5, 0.8)	176	1.8%	(1.5, 2.2)	149	0.4%	(0.3, 0.5)
	Non-flavored	21	26.5%	(17.7, 37.5)	80	45.8%	(39.1, 52.6)	79	53.0%	(41.9, 63.8)
	Any flavor use	48	73.6%	(62.5, 82.3)	96	54.2%	(47.4, 60.9)	70	47.0%	(36.2, 58.1)	g,h*
	Menthol or Mint	-	-	-	80	86.2%	(77.9, 91.7)	-		-
	Clove or Spice	-	-	-	-	-	-	-	-	-
	Fruit	-	-	-	13	11.0%	(6.0, 19.1)	-	-	-
	Chocolate	-	-	-	-	-	-	-	-	-
	Alcoholic Drink	-	-	-	-	-	-	-	-	-
	Candy/Other Sweets	-	-	-	-	-	-	-	-	-
	Other Flavor	-	-	-	-	-	-	-	-	-

^†Age groups (youth, young adults, and adults 25+) were categorized based on age at Wave 2.

^‡Respondents could report using multiple products.

^-Dashes represent estimates that have been suppressed because they are statistically unreliable. The estimates are based on a denominator sample size of less than 50, or the relative standard error (RSE) of the estimate or its complement is greater than 30%.

^§Percentages are weighted to represent the U.S. population and CIs are estimated using the balanced repeated replication (BRR) method. Flavored pipe tobacco and dissolvable tobacco use were not assessed in youth. Results for dissolvable tobacco among adults are not presented due to small sample size, n < 50.

^aFor any tobacco product use, new product users started to use any new product between Waves 1 and 2. This can include never users at Wave 1 who started tobacco use at Wave 2 and ever users at Wave 1 who reported use of a new product(s) at Wave 2. “Don’t know” and refused to answer for any part of the definition of ever use or any flavor use (see below) were excluded from the denominator. For product-specific use, new product users reported using the specific product for the first time between Waves 1 and 2. Unweighted n’s and unweighted percentages are presented for the following age groups (differing from weighted estimates provided in the table): Among the full sample of youth (n=12,172), n=1,441 (11.8%) were new product users, n=8,408 (69.1%) reported no new product use, and n=2,323 (19.1%) did not have information on new use between Wave 1 and Wave 2 (n=2,091 of these instances reflect aged-up youth who were not interviewed in Wave 1 and the remainder having missing data). Among the full sample of young adults (n=8,174), n=2,526 (30.9%) were new product users, n=5,611 (68.6%) reported no new product use, and n=37 (0.5%) did not have information on new use between Wave 1 and Wave 2. Among the full sample of adults (25+) (n=20,183), n=2,807 (13.9%) were new product users, n=17,243 (85.4%) reported no new product use, and n=133 (0.7%) did not provide information on new use between Wave 1 and Wave 2.

^bIn the full sample of youth, weighted percentage of new product users between Wave 1 and Wave 2.

^cIn the full sample of young adult, weighted percentage of new product users between Wave 1 and Wave 2.

^dIn the full sample of adults (25+), weighted percentage of new product users between Wave 1 and Wave 2.

^eFor any tobacco product, any flavor use is defined as any new product use that was “flavored to taste like menthol, mint, clove, spice, candy, fruit, chocolate, alcohol (such as wine or cognac), or other sweets”, among new product users (denominator). “No” or “I don’t know” was categorized as non-flavored use. “Don’t know” or refused to answer whether the first product was flavored were excluded from the denominator. Product-specific any flavor use is defined as using a first new product that was “flavored to taste like menthol, mint, clove, spice, candy, fruit, chocolate, alcohol (such as wine or cognac), or other sweets”, among New Product-specific Users (denominator). Any flavor use among new product users of pipe and dissolvable tobacco was not assessed in youth. For youth new product users (unweighted n, unweighted %) (n=1,441), n=79 (5.5%) did not report whether they had any flavor use between Wave 1 and Wave 2. For young adult new product users (n=2,526), n=70 (2.8%) did not report whether they had any flavor use between Wave 1 and Wave 2. For adult new product users (n=2,807), n=62 (2.2%) did not report whether they had any flavor use between Wave 1 and Wave 2.

^fFor any tobacco product, first [flavor type] use (e.g., first menthol use), is defined as any new use of any product of that flavor type between Waves 1 and 2, among any flavor users (denominator). Product-specific first [flavor type] use is defined as any new use of that specific flavor and product between Waves 1 and 2, among product-specific any flavor users (denominator). Any flavor users could select all flavor types that applied.

^gProportion of youth is significantly different from proportion of young adults at p<0.05

^hProportion of youth is significantly different from proportion of adults (25+) at p<0.05

ⁱProportion of young adults is significantly different from proportion of adults (25+) at p<0.05

^*Overall Pearson’s chi-square test is significant at the p<0.05 level

^jYouth are not asked questions about use of flavored pipe tobacco.

Source: Population Assessment of Tobacco and Health (PATH) Study, Wave 2

Table 2.

Use of a flavored tobacco product and flavor of product used in the past 30-days among flavored tobacco users by age group^† and tobacco product^‡ in Wave 2

		Past 30-day use at Wave 2 and Past 30-day Flavor Type among Past 30-day Flavored Tobacco Users§
		Youth (12–17)			Young Adults (18–24)			Adults (25+)
		Frequency	Percent	95% CI	Frequency	Percent	95% CI	Frequency	Percent	95% CI	Hypothesis Test
Any Tobacco Product	Past 30-Day Usersa	920	7.8%b	(7.2, 8.4)	3726	38.2%c	(36.8, 39.7)	10346	25.7%d	(25.1, 26.3)
	Non-flavorede	262	27.7%	(24.9, 30.6)	1186	32.4%	(30.6, 34.3)	5538	55.2%	(53.8, 56.6)
	Any flavor usee	658	72.3%	(69.4, 75.1)	2540	67.6%	(65.7, 69.4)	4808	44.7%	(43.3, 46.2)	g,h,i*
	Menthol or Mintf	340	51.7%	(47.8, 55.6)	1742	68.1%	(66.1, 70.0)	3768	76.1%	(74.6, 77.5)	g,h,i*
	Clove or Spicef	33	5.0%	(3.6, 7.1)	136	5.0%	(4.3, 5.8)	144	3.0%	(2.5, 3.5)	h,i*
	Fruitf	353	54.4%	(50.7, 57.9)	1234	49.6%	(47.1, 52.1)	1068	22.9%	(21.6, 24.3)	g,h,i*
	Chocolatef	60	8.9%	(7.0, 11.2)	191	6.9%	(6.0, 8.0)	208	4.1%	(3.5, 4.8)	h,i*
	Alcoholic Drinkf	61	9.7%	(7.5, 12.4)	310	11.9%	(10.7, 13.2)	248	5.0%	(4.3, 5.8)	h,i*
	Candy/Other Sweetsf	229	35.3%	(31.3, 39.5)	707	27.3%	(25.4, 29.2)	627	13.3%	(12.2, 14.4)	g,h,i*
	Other Flavorf	149	22.5%	(18.5, 26.9)	363	14.1%	(12.6, 15.8)	496	10.5%	(9.4, 11.8)	g,h,i*
Cigarettes	Past 30-Day Users	480	3.9%	(3.5, 4.4)	2573	25.8%	(24.7, 27.0)	8565	20.6%	(20.0, 21.2)
	Non-flavored	226	47.1%	(42.4, 51.8)	1211	48.0%	(45.7, 50.4)	5369	64.4%	(63.0, 65.6)
	Any flavor use	254	52.9%	(48.2, 57.6)	1362	52.0%	(49.7, 54.3)	3196	35.7%	(34.4, 37.0)	h,i*
	Menthol or Mint	222	88.1%	(83.1, 91.7)	1311	96.2%	(94.8, 97.2)	3110	96.6%	(95.6, 97.4)	g,h*
	Clove or Spice	-	-	-	50	3.4%	(2.6, 4.4)	51	1.4%	(1.1, 1.9)	i
	Fruit	29	10.8%	(7.5, 15.3)	67	4.6%	(3.5, 6.0)	63	1.7%	(1.3, 2.3)	g,h,i*
	Chocolate	9	3.3%	(1.9, 5.7)	25	1.8%	(1.1, 2.9)	20	0.5%	(0.4, 0.8)	h,i*
	Alcoholic Drink	11	4.2%	(2.4, 7.5)	41	2.9%	(2.0, 4.1)	49	1.7%	(1.2, 2.5)
	Candy/Other Sweets	19	7.8%	(5.1, 11.7)	42	3.1%	(2.2, 4.2)	44	1.3%	(0.9, 2.0)	g,h,i*
	Other Flavor	31	11.7%	(7.8, 17.0)	46	2.9%	(2.2, 3.9)	57	2.1%	(1.5, 2.9)	g,h*
E-cigarettes	Past 30-Day Users	414	3.6%	(3.2, 4.1)	960	9.7%	(8.8, 10.6)	1907	4.6%	(4.3, 4.9)
	Non-flavored	85	20.6%	(16.6, 25.4)	170	16.8%	(14.1, 19.8)	670	35.7%	(33.1, 38.4)
	Any flavor use	329	79.4%	(74.7, 83.4)	790	83.3%	(80.2, 85.9)	1237	64.3%	(61.6, 66.9)	h,i*
	Menthol or Mint	66	20.4%	(16.6, 25.0)	248	30.9%	(27.1, 34.9)	515	40.7%	(37.8, 43.8)	g,h,i*
	Clove or Spice	16	5.1%	(3.0, 8.7)	17	2.6%	(1.5, 4.2)	41	2.9%	(2.2, 3.8)
	Fruit	231	71.0%	(65.5, 75.9)	532	68.7%	(64.8, 72.3)	572	46.3%	(43.2, 49.5)	h,i*
	Chocolate	34	10.4%	(7.6, 14.0)	82	9.8%	(7.7, 12.4)	88	6.6%	(5.3, 8.3)	h,i*
	Alcoholic Drink	25	8.4%	(5.7, 12.1)	66	8.0%	(6.0, 10.5)	54	4.1%	(3.1, 5.3)	h,i*
	Candy/Other Sweets	160	49.1%	(42.4, 55.9)	380	48.5%	(44.8, 52.3)	398	33.7%	(30.7, 36.8)	h,i*
	Other Flavor	77	23.1%	(18.3, 28.7)	157	19.8%	(16.8, 23.1)	240	19.0%	(16.4, 21.9)
Traditional Cigars	Past 30-Day Users	74	0.6%	(0.5, 0.8)	397	4.0%	(3.5, 4.6)	831	2.3%	(2.1, 2.5)
	Non-flavored	43	58.4%	(46.0, 69.9)	295	76.2%	(70.7, 81.0)	605	75.9%	(72.7, 78.8)
	Any flavor use	31	41.6%	(30.1, 54.0)	102	23.8%	(19.0, 29.3)	226	24.1%	(21.2, 27.3)	g,h
	Menthol or Mint	-	-	-	27	25.2%	(17.5, 34.9)	69	29.0%	(22.9, 36.1)
	Clove or Spice	-	-	-	-	-	-	-	-	-
	Fruit	-	-	-	41	39.5%	(30.4, 49.3)	73	33.1%	(26.2, 40.9)	g,h
	Chocolate	-	-	-	19	16.5%	(10.5, 24.9)	38	15.7%	(11.5, 21.2)
	Alcoholic Drink	-	-	-	23	21.9%	(14.6, 31.6)	42	19.3%	(13.9, 26.1)
	Candy/Other Sweets	-	-	-	30	29.2%	(20.4, 40.0)	46	19.2%	(14.1, 25.8)
	Other Flavor	-	-	-	19	17.7%	(11.0, 27.2)	42	19.2%	(14.0, 25.6)
Traditional Cigars	Past 30-Day Users	74	0.6%	(0.5, 0.8)	397	4.0%	(3.5, 4.6)	831	2.3%	(2.1, 2.5)
	Non-flavored	43	58.4%	(46.0, 69.9)	295	76.2%	(70.7, 81.0)	605	75.9%	(72.7, 78.8)
	Any flavor use	31	41.6%	(30.1, 54.0)	102	23.8%	(19.0, 29.3)	226	24.1%	(21.2, 27.3)	g,h
	Menthol or Mint	-	-	-	27	25.2%	(17.5, 34.9)	69	29.0%	(22.9, 36.1)
	Clove or Spice	-	-	-	-	-	-	-	-	-
	Fruit	-	-	-	41	39.5%	(30.4, 49.3)	73	33.1%	(26.2, 40.9)	g,h
	Chocolate	-	-	-	19	16.5%	(10.5, 24.9)	38	15.7%	(11.5, 21.2)
	Alcoholic Drink	-	-	-	23	21.9%	(14.6, 31.6)	42	19.3%	(13.9, 26.1)
	Candy/Other Sweets	-	-	-	30	29.2%	(20.4, 40.0)	46	19.2%	(14.1, 25.8)
	Other Flavor	-	-	-	19	17.7%	(11.0, 27.2)	42	19.2%	(14.0, 25.6)
Cigarillos	Past 30-Day Users	121	1.0%	(0.8, 1.2)	628	5.9%	(5.5, 6.4)	807	2.0%	(1.8, 2.1)
	Non-flavored	59	47.7%	(38.6, 57.0)	301	49.0%	(44.3, 53.7)	388	48.9%	(44.8, 53.0)
	Any flavor use	62	52.3%	(43.0, 61.4)	327	51.0%	(46.3, 55.7)	419	51.1%	(47.0, 55.2)
	Menthol or Mint	-	-	-	41	12.5%	(8.9, 17.3)	108	24.1%	(19.3, 29.7)	i
	Clove or Spice	-	-	-	-	-	-	26	5.9%	(3.9, 8.7)
	Fruit	41	66.3%	(53.9, 76.8)	182	55.4%	(49.2, 61.5)	193	48.5%	(41.6, 55.4)	h
	Chocolate	-	-	-	29	8.4%	(5.9, 12.0)	52	12.0%	(8.5, 16.7)
	Alcoholic Drink	13	20.9%	(12.9, 32.1)	103	32.7%	(26.5, 39.7)	93	22.4%	(18.1, 27.4)
	Candy/Other Sweets	20	36.8%	(24.3, 51.3)	79	23.8%	(18.9, 29.5)	86	21.6%	(16.6, 27.6)
	Other Flavor	-	-	-	53	16.3%	(11.9, 21.9)	57	12.6%	(9.7, 16.3)
Filtered Cigars	Past 30-Day Users	49	0.4%	(0.3, 0.6)	245	2.2%	(1.9, 2.5)	536	1.3%	(1.2, 1.5)
	Non-flavored	-	-	-	128	52.8%	(45.3, 60.1)	252	49.8%	(45.2, 54.5)
	Any flavor use	-	-	-	117	47.2%	(39.9, 54.7)	284	50.2%	(45.5, 54.8)
	Menthol or Mint	-	-	-	39	33.6%	(24.9, 43.6)	136	44.2%	(37.8, 50.7)
	Clove or Spice	-	-	-	24	21.8%	(14.7, 31.2)	26	10.3%	(6.2, 16.5)	i
	Fruit	-	-	-	52	43.7%	(33.4, 54.6)	90	31.9%	(26.2, 38.1)
	Chocolate	-	-	-	-	-	-	-	-	-
	Alcoholic Drink	-	-	-	16	14.0%	(8.3, 22.5)	30	13.6%	(8.5, 21.0)
	Candy/Other Sweets	-	-	-	14	13.2%	(7.7, 21.5)	41	14.0%	(10.1, 19.3)
	Other Flavor	-	-	-	-	-	-	-	-	-
Any Cigarsj	Past 30-Day Users	180	1.5%	(1.3, 1.8)	957	9.4%	(8.7, 10.1)	1641	4.2%	(4.0, 4.5)
	Non-flavored	86	46.3%	(38.9, 53.9)	508	55.5%	(51.6, 59.3)	940	60.4%	(57.5, 63.1)
	Any flavor use	94	53.7%	(46.1, 61.1)	449	44.5%	(40.7, 48.4)	701	39.6%	(36.9, 42.5)	i
	Menthol or Mint	15	16.9%	(10.3, 26.5)	85	18.7%	(15.1, 22.9)	244	32.6%	(28.5, 37.1)	h,i
	Clove or Spice	-	-	-	39	8.8%	(6.5, 11.8)	52	8.0%	(5.8, 11.1)
	Fruit	65	68.3%	(58.2, 77.0)	241	52.9%	(47.4, 58.4)	275	41.0%	(35.7, 46.4)	g,h,i*
	Chocolate	11	10.3%	(5.8, 17.6)	47	9.4%	(7.1, 12.2)	83	11.7%	(9.2, 14.9)
	Alcoholic Drink	16	17.4%	(11.2, 26.1)	126	28.0%	(23.4, 33.2)	129	18.5%	(15.4, 22.2)	g,i*
	Candy/Other Sweets	26	30.0%	(21.2, 40.5)	104	22.5%	(18.6, 27.0)	136	19.8%	(16.3, 23.8)
	Other Flavor	17	18.2%	(10.5, 29.6)	70	15.5%	(11.8, 20.0)	97	13.6%	(11.2, 16.5)
Pipe	Past 30-Day Users	N/A			108	1.1%	(0.9, 1.4)	183	0.4%	(0.4, 0.5)
	Non-flavored	N/A			75	72.4%	(61.9, 80.9)	134	72.5%	(64.2, 79.6)
	Any flavor use	N/A			33	27.6%	(19.1, 38.1)	49	27.5%	(20.5, 35.8)
	Menthol or Mint	N/A			14	41.1%	(22.9, 62.1)	15	30.1%	(16.4, 48.5)
	Clove or Spice	N/A			-	-	-	-	-	-
	Fruit	N/A			-	-	-	17	30.9%	(18.1, 47.5)
	Chocolate	N/A			-	-	-	-	-	-
	Alcoholic Drink	N/A			-	-	-	-	-	-
	Candy/Other Sweets	N/A			-	-	-	-	-	-
	Other Flavor	N/A			-	-	-	-	-	-
Hookah	Past 30-Day Users	153	1.2%	(1.1, 1.5)	1113	11.7%	(10.9, 12.5)	488	1.2%	(1.1, 1.4)
	Non-flavored	45	28.5%	(21.9, 36.3)	368	32.0%	(29.0, 35.2)	179	37.8%	(32.6, 43.3)
	Any flavor use	108	71.5%	(63.7, 78.1)	745	68.0%	(64.9, 71.0)	309	62.2%	(56.7, 67.4)
	Menthol or Mint	34	32.5%	(24.4, 41.8)	266	36.9%	(32.6, 41.3)	103	33.0%	(26.5, 40.2)
	Clove or Spice	-	-	-	41	4.7%	(3.5, 6.3)	22	7.2%	(4.7, 11.0)
	Fruit	86	79.8%	(70.8, 86.5)	612	81.8%	(78.1, 85.1)	230	75.7%	(69.8, 80.8)
	Chocolate	-	-	-	70	8.4%	(6.5, 10.6)	35	10.2%	(7.2, 14.3)
	Alcoholic Drink	14	12.3%	(6.8, 21.2)	115	14.7%	(12.2, 17.6)	32	7.8%	(5.3, 11.2)	i*
	Candy/Other Sweets	49	45.4%	(35.5, 55.7)	296	35.9%	(31.9, 40.1)	91	26.1%	(20.7, 32.3)	h,i*
	Other Flavor	13	11.1%	(6.5, 18.4)	88	11.7%	(8.9, 15.2)	39	9.6%	(6.5, 13.8)
Smokeless Tobacco	Past 30-Day Users	123	1.1%	(0.9, 1.4)	419	4.3%	(3.9, 4.7)	952	2.5%	(2.3, 2.8)
	Non-flavored	47	37.3%	(29.7, 45.7)	181	42.8%	(37.2, 48.5)	480	51.8%	(46.3, 57.3)
	Any flavor use	76	62.7%	(54.3, 70.3)	238	57.2%	(51.5, 62.8)	472	48.2%	(42.7, 53.7)	h,i*
	Menthol or Mint	56	73.1%	(62.8, 81.4)	184	78.2%	(71.5, 83.6)	371	76.7%	(70.9, 81.7)
	Clove or Spice	-	-	-	-	-	-	-	-	-
	Fruit	14	19.5%	(12.0, 29.9)	33	13.1%	(9.6, 17.7)	42	8.6%	(5.8, 12.5)
	Chocolate	-	-	-	-	-	-	-	-	-
	Alcoholic Drink	-	-	-	-	-	-	-	-	-
	Candy/Other Sweets	-	-	-	-	-	-	-	-	-
	Other Flavor	24	33.1%	(23.8, 44.0)	45	18.2%	(13.8, 23.7)	79	18.2%	(14.3, 22.8)	g,h*
Snus	Past 30-Day Users	39	0.4%	(0.3, 0.5)	131	1.3%	(1.1, 1.6)	148	0.4%	(0.3, 0.5)
	Non-flavored	-	-	-	41	29.9%	(21.4, 40.1)	41	27.3%	(20.6, 35.1)
	Any flavor use	-	-	-	90	70.1%	(59.9, 78.6)	107	72.7%	(64.9, 79.4)
	Menthol or Mint	-	-	-	78	88.5%	(80.2, 93.6)	-	-	-
	Clove or Spice	-	-	-	-	-	-	-	-	-
	Fruit	-	-	-	-	-	-	-	-	-
	Chocolate	-	-	-	-	-	-	-	-	-
	Alcoholic Drink	-	-	-	-	-	-	-	-	-
	Candy/Other Sweets	-	-	-	-	-	-	-	-	-
	Other Flavor	-	-	-	-	-	-	-	-	-

^†Age groups (youth, young adults, and adults 25+) were categorized based on age at Wave 2.

^‡Respondents could report multiple products.

^-Dashes represent estimates that have been suppressed because they are statistically unreliable. The estimates are based on a denominator sample size of less than 50, or relative standard error (RSE) of the estimate or its complement is greater than 30%.

^§Percentages are weighted to represent the U.S. population and CIs are estimated using the balanced repeated replication (BRR) method. Flavored pipe tobacco and dissolvable tobacco use were not assessed in youth. Results for dissolvable tobacco are not presented due to small sample size, n < 50.

^aFor overall tobacco, past 30-day users are users of any product in the past 30 days. For product-specific use, past 30-day users used the product (even one or two times) in the past 30 days. “Don’t know” or refused to answer were excluded from the denominator. Unweighted n’s and percentages are presented for the following age groups: Among the full sample of youth (n=12,172), n=929 (7.6%) were past 30-day users, n=10,867 (89.3%) reported no past 30-day use, and n=376 (3.1%) did not provide information on past 30-day use in Wave 2. For youth past 30-day users (n=929), n=9 (1.0%) did not report whether they had any flavor use in the past 30 days. Among the full sample of young adults (n=8,174), n=4,054 (49.6%) were past 30-day users, n=4,062 (49.7%) reported no past 30-day use, and n=58 (0.7%) did not provide information on past 30-day use in Wave 2. For young adult past 30-day users (n=4,054), n=328 (8.1%) did not report whether they had any flavor use in the past 30 days. Among the full sample of adults (25+) (n=20,183), n=10,793 (53.5%) were past 30-day users, n=9,280 (46.0%) reported no past 30-day use, and n=110 (0.6%) did not provide information on past 30-day use in Wave 2. For adult past 30-day users (n=10,793), n=447 (4.1%) did not report whether they had any flavor use in the past 30 days.

^bIn the full sample of youth, weighted percentage of past 30-day users with any flavor use in the past 30 days.

^cIn the full sample of young adult, weighted percentage of past 30-day users with any flavor use in the past 30 days.

^dIn the full sample of adults (25+), weighted percentage of past 30-day users with any flavor use in the past 30 days.

^eFor overall tobacco, any flavor use is defined as using any flavored product in the past 30 days “flavored to taste like menthol, mint, clove, spice, candy, fruit, chocolate, alcohol (such as wine or cognac), or other sweets”, among past 30-day users (denominator). For product-specific use, any flavor use is defined as using that specific product in the past 30 days “flavored to taste like menthol, mint, clove, spice, candy, fruit, chocolate, alcohol (such as wine or cognac), or other sweets”, among past 30-day users (denominator). Any flavor use among past 30-day users of pipe and dissolvable tobacco was not assessed in youth.

^fFor overall tobacco, past 30-day [flavor type] (e.g., past 30-day menthol use) is defined as using any product of that flavor type in the past 30 days, among any flavor use (denominator). For product-specific use, past 30-day [flavor type] is defined as using that product of that flavor type in the past 30 days, among any flavor use (denominator). Any flavor users could select all flavor types that applied. Past 30-day flavor type use of pipe and dissolvable tobacco was not assessed in youth.

^gProportion of youth is significantly different from proportion of young adults at p<0.05

^hProportion of youth is significantly different from proportion of adults (25+) at p<0.05

ⁱProportion of young adults is significantly different from proportion of adults (25+) at p<0.05

^jAny cigar use is comprised of any use of traditional cigars, cigarillos, or filtered cigars in the past 30 days.

^*Overall Pearson’s chi-square test is significant at the p<0.05 level

Source: Population Assessment of Tobacco and Health (PATH) Study, Wave 2

Descriptive analyses of adults’ (18+) self-reported flavor of their brand were weighted and examined for consistency of response among current and recent former established users who reported both a brand and flavor type. The top 10 expert-coded primary flavors (including non-flavored like “Marlboro non-menthol” or ambiguous flavors like “Swisher Sweets unknown”) were listed by prevalence of use for each tobacco product type. Table 3 shows the weighted percentage of self-reported brand flavor type for each expert-coded primary flavor by product.. Finally, Table S4 shows the unweighted concordance of brand flavor type as identified by respondents and the expert panel, reporting percent agreement and kappa reliability estimates. “Substantial” concordance was benchmarked at κ > 0.6 and “moderate” concordance at κ between 0.4 and 0.6.³⁵ For all analyses, data with denominators less than 50 or relative standard error of an estimate or its complement of more than 30% were suppressed.³⁶

Table 3.

Adult Established Users (age 18+) characterization of the flavor of their brand compared with primary flavor based on expert coding (Weighted %)

					Percentage of Usersa Self-Reporting Brand Flavor Use Statusb			Self-reported brand flavor typec
Product		Expert-Coded Primary Flavord	Weighted %e	Sample size (n)	“I don’t know”	Non-flavored	Any Flavor	Mint/ menthol	Clove/ Spice	Fruit	Chocolate	Alcohol	Candy or Sweet	Some other flavor
Manufactured Cigarettes (n=10,055)	1125	Marlboro Non-menthol	31.0%	2957	2.3%	95.7%	2.0%	69.4%	-f	-	-	-	-	-
	1144	Newport Mentholg	14.2%	1605	4.0%	11.8%	84.1%	99.7%	-	-	-	-	-	-
	1033	Camel Non-menthol	7.9%	793	4.6%	92.0%	3.4%	-	-	-	-	-	-	-
	1125	Marlboro Menthol	6.8%	741	-	2.3%	96.3%	100.0%	-	-	-	-	-	-
	1033	Camel Menthol	5.4%	644	-	8.9%	88.9%	100.0%	-	-	-	-	-	-
	1152	Pall Mall Non-menthol	4.7%	441	-	97.5%	-	-	-	-	-	-	-	-
	1012	American Spirit Non-menthol	2.8%	294	-	94.5%	-	-	-	-	-	-	-	-
	1113	Kool Menthol	2.3%	211	-	15.4%	80.9%	98.3%	-	-	-	-	-	-
	1152	Pall Mall Menthol	1.9%	195	-	-	97.4%	100.0%	-	-	-	-	-	-
	1115	L & M Non-menthol	1.8%	185	-	98.6%	-	-	-	-	-	-	-	-
Traditional Cigars (n=849)	1047	Cohiba Non-flavored	12.1%	82	-	76.3%	-	-	-	-	-	-	-	-
	1204	Swisher Sweets unknownh	6.7%	71	-	68.9%	25.6%	-	-	-	-	-	-	-
	1017	Arturo Fuente Non-flavored	6.2%	56	-	93.1%	-	-	-	-	-	-	-	-
	1061	Dutch Masters Non-flavored	6.1%	46	-	-	-	-	-	-	-	-	-	-
	1124	Macanudo Non-flavored	6.0%	52	-	89.6%	-	-	-	-	-	-	-	-
	1174	Romeo y Julieta Non-flavored	4.8%	34	-	-	-	-	-	-	-	-	-	-
	1061	Dutch Masters Fruit	-	23	-	-	-	-	-	-	-	-	-	-
	1084	Garcia y Vega unknown	3.8%	30	-	-	-	-	-	-	-	-	-	-
	1204	Swisher Sweets Fruit	3.5%	38	-	-	-	-	-	-	-	-	-	-
	1231	White Owl Fruit	3.1%	32	-	-	-	-	-	-	-	-	-	-
Cigarillos (n=1338)	1024	Black & Mild Non-flavored	20.0%	269	5.0%	72.1%	22.9%	51.2%	-	15.1%	-	-	-	-
	1024	Black & Mild Alcohol	17.0%	258	-	17.4%	79.0%	6.4%	-	16.4%	-	83.1%	-	5.4%
	1204	Swisher Sweets Fruit	13.9%	174	-	16.0%	77.6%	14.2%	-	80.4%	-	-	14.9%	-
	1204	Swisher Sweets unknown	12.9%	146	-	65.9%	19.6%	-	-	-	-	-	-	-
	1024	Black & Mild Candy/Other Sweets	7.4%	94	-	31.8%	63.0%	-	-	40.6%	-	-	25.0%	-
	1231	White Owl Fruit	4.7%	82	-	18.0%	73.5%	-	-	73.4%	-	-	19.9%	-
	1024	Black & Mild unknown	2.1%	36	-	-	-	-	-	-	-	-	-	-
	1231	White Owl unknown	1.9%	29	-	-	-	-	-	-	-	-	-	-
	1156	Phillies Non-flavored	-	12	-	-	-	-	-	-	-	-	-	-
	1239	Zig Zag Fruit	-	18	-	-	-	-	-	-	-	-	-	-
Filtered Cigars (n=632)	1204	Swisher Sweets Fruit	18.2%	129	-	17.0%	77.3%	-	-	75.6%	-	-	17.3%	-
	1204	Swisher Sweets Non-flavored	13.8%	82	-	72.9%	21.2%	-	-	-	-	-	-	-
	1156	Phillies Non-flavored	10.5%	58	-	57.4%	34.2%	-	-	-	-	-	-	-
	1242	Prime Time Fruit	8.1%	64	-	-	89.3%	-	-	90.7%	-	-	14.0%	-
	1042	Cheyenne Non-flavored	5.5%	34	-	-	-	-	-	-	-	-	-	-
	1002	305’s Non-flavored	4.3%	24	-	-	-	-	-	-	-	-	-	-
	1042	Cheyenne Fruit	4.3%	33	-	-	-	-	-	-	-	-	-	-
	1156	Phillies Candy/Other Sweets	3.5%	23	-	-	-	-	-	-	-	-	-	-
	1042	Cheyenne Menthol	3.1%	20	-	-	-	-	-	-	-	-	-	-
	1182	Santa Fe Non-flavored	-	9	-	-	-	-	-	-	-	-	-	-
Pipe (n=325)	1161	Prince Albert Non-flavored	15.7%	34	-	-	-	-	-	-	-	-	-	-
	1037	Captain Black Non-flavored	13.3%	49	-	-	-	-	-	-	-	-	-	-
	1094	Half & Half Non-flavored	10.8%	20	-	-	-	-	-	-	-	-	-	-
	1189	Sir Walter Raleigh Non-flavored	-	15	-	-	-	-	-	-	-	-	-	-
	1029	Borkum Riff Non-flavored	-	12	-	-	-	-	-	-	-	-	-	-
	1037	Captain Black Fruit	4.7%	17	-	-	-	-	-	-	-	-	-	-
	1081	Gambler Non-flavored	-	20	-	-	-	-	-	-	-	-	-	-
	1029	Borkum Riff Fruit	3.0%	8	-	-	-	-	-	-	-	-	-	-
	1161	Prince Albert Candy/Other Sweet	2.4%	3	-	-	-	-	-	-	-	-	-	-
	1061	Dutch Masters Fruit	-	11	-	-	-	-	-	-	-	-	-	-
Hookah (n=1,920)	1198	Starbuzz Fruit	19.8%	396	8.4%	12.6%	79.0%	28.2%	-	80.4%	-	6.6%	27.0%	7.3%
	1073	Fantasia Fruit	7.1%	141	-	13.9%	82.1%	14.0%	-	76.4%	-	11.5%	28.2%	14.8%
	1095	Havana Fruit	7.0%	136	16.3%	17.3%	66.4%	18.1%	-	84.4%	-	-	15.5%	-
	1007	Al Amir Fruit	5.6%	90	8.7%	36.1%	55.3%	17.0	-	77.4%	-	-	22.0%	-
	1253	Al Fakher Fruit	5.5%	91	-	14.8%	81.2%	19.0%	-	86.0%	-	-	-	-
	1100	Hydro Herbal Fruit	4.4%	78	17.2%	27.3%	55.6%	-	-	-	-	-	-	-
	1098	HookahFina Fruit	4.3%	88	22.4%	16.2%	61.5%	21.1%	-	79.9%	-	-	-	-
	1195	Social Smoke Fruit	3.7%	72	19.0%	-	69.4%	-	-	-	-	-	-	-
	1101	Inhale Fruit	3.3%	72	25.0%	33.5%	41.4%	-	-	-	-	-	-	-
	1073	Fantasia Alcohol	3.1%	67	-	16.7%	77.1%	-	-	65.5%	-	41.9%	31.5%	-
Smokeless Tobacco (n=1,576)	1049	Copenhagen Non-flavored	23.2%	359	2.6%	86.4%	11.0%	-	-	-	-	-	-	-
	1092	Grizzly Menthol	19.3%	343	-	12.7%	83.1%	85.2%	-	-	-	-	-	15.8%
	1092	Grizzly Non-flavored	9.0%	154	-	71.4%	26.2%	-	-	-	-	-	-	-
	1049	Copenhagen Menthol	7.4%	140	-	16.9%	79.2%	88.2%	-	-	-	-	-	12.1%
	1190	Skoal Menthol	6.1%	90	-	-	93.4%	95.9%	-	-	-	-	-	-
	1190	Skoal Non-flavored	5.3%	69	-	54.9%	39.2%	-	-	-	-	-	-	-
	1112	Kodiak Menthol	4.5%	54	-	-	74.3%	-	-	-	-	-	-	-
	1190	Skoal Fruit	3.4%	51	-	-	92.4%	-	-	-	-	-	-	-
	1170	Red Man Non-flavored	3.2%	43	-	-	-	-	-	-	-	-	-	-
	1240	Beechnut Non-flavored	3.0%	33	-	-	-	-	-	-	-	-	-	-
Snus (n=342)	1033	Camel Menthol	47.3%	166	-	6.1%	88.9%	92.5%	-	-	-	-	-	-
	1190	Skoal Menthol	27.5%	84	-	-	86.5%	90.6%	-	-	-	-	-	-
	1033	Camel Non-flavored	10.6%	38	-	-	-	-	-	-	-	-	-	-
	1125	Marlboro Menthol	3.6%	16	-	-	-	-	-	-	-	-	-	-
	1085	General Menthol	-	10	-	-	-	-	-	-	-	-	-	-
	1125	Marlboro Non-flavored	-	8	-	-	-	-	-	-	-	-	-	-
	1190	Skoal unknown	-	8	-	-	-	-	-	-	-	-	-	-
	1190	Skoal Fruit	-	3	-	-	-	-	-	-	-	-	-	-
	1085	General Non-flavored	-	4	-	-	-	-	-	-	-	-	-	-
	1190	Skoal Non-flavored	-	2	-	-	-	-	-	-	-	-	-	-

^aAnalyses of users are restricted to current established and recent former established adult smokers/users who reported a brand/sub-brand and self-reported flavor use and flavor type of their brand. For hookah, analyses of users are restricted to current established, recent former established, current experimental, and past 30-day users who reported a brand/sub-brand and self-reported flavor use and flavor type of their brand.

^bDenominator of flavor status is the row sample size (weighted)

^cDenominator is the number of respondents indicating their product was flavored (any flavor; weighted)

^dTop ten expert-coded primary-flavors listed for each product by frequency of use at Wave 2. Expert-coded primary flavors are based on expert panel coding of the primary flavor of the respondents’ brand/sub-brand.

^eWeighted frequency of specific brand among current established and recent former established users of the specific product

^fDashes (−) represent estimates that have been suppressed because they are statistically unreliable. The estimates are based on a denominator sample size of less than 50, or the relative standard error (RSE) of the estimate or its complement is greater than 30%.

^gRed Bold Underline indicates “flavored” brands

^hBlue bold Italics indicates brands with “unknown” flavor status

Results

First FTP Use

Between Waves 1 and 2, 11.4% of youth, 26.9% of YAs, and 8.1% of adults 25+ used a new tobacco product. Of those, 74.3% of youth, 59.5% of YAs, and 45.1% of adults used at least one FTP at first use (Table 1). Among new product users of any age, e-cigarette, hookah, and snus users most commonly used an FTP.

Flavor Type Used at First Use Among New FTP Users

Among youth and YA new FTP users, fruit was the most prevalent flavor at first use, with the highest use among YAs (60.4%), followed by youth (56.7%), both differing significantly from adult 25+ FTP users (39.2%) (Table 1 and Figure 1, Any Product). A significantly greater prevalence of adult 25+ new FTP users used menthol/mint flavor overall (44.6%) at first use, compared with youth (30.4%) or YAs (39.6%). The prevalence of any first FTP use of candy/sweet flavor was significantly lower for adult 25+ new FTP users (18.8%), compared with youth (27.5%) or YAs (30.9%). Alcoholic drink flavor at first FTP use had the highest prevalence among YA new FTP users (9.9%) and adults 25+ (7.2%), compared with youth (4.3%). All age groups used clove/spice less often than other flavors.

Figure 1.

First flavour type among new flavoured product users by age group.

Product-Specific First FTP Type Among New Tobacco Product Users

Table S2 includes tobacco product types with at least 50 (unweighted) new users and estimates for at least one flavor type across ages and products (see Table 1 for prevalence among FTP users).

Menthol/mint.

For new smokeless tobacco and snus users, youth had higher prevalence of menthol/mint at first use than adults 25+. For e-cigarettes and any cigar, adults 25+ had higher prevalence of menthol/mint flavor at first use than youth.

Fruit.

Youth had significantly higher prevalence of first use of fruit-flavored e-cigarettes, cigarillos, filtered cigars, any cigar and hookah, compared with adults 25+.

Candy/sweet.

Adults 25+ had a lower prevalence of candy/sweet e-cigarette and hookah use than other ages at first use.

Past 30-Day Flavor Use

Table 2 shows prevalence of FTP use among past 30-day tobacco users and flavor type among past 30-day FTP users. At Wave 2, among past 30 day tobacco product users, 72.3% of youth, 67.6% of YAs, and 44.7% of adults 25+ had used at least one FTP. Across products for each age group, e-cigarette, hookah, and snus users most commonly used an FTP in the past 30 days.

Any Tobacco Past 30-Day Flavor Type

An inverse age gradient for fruit, candy/sweet, and other flavor use was evident among past 30-day FTP users (see Table 2 and Figure 2, Any Tobacco), with highest use among youth. Menthol/mint product use patterns differed, with the highest prevalence of use among adults 25+ (76.1%), followed by YA (68.1%) and youth FTP users (51.7%). Prevalence of past 30-day use of chocolate, alcohol, or clove/spice flavors was significantly higher among youth and YA FTP users compared with adults 25+.

Figure 2.

Past 30-day flavour type among past 30-day flavoured tobacco users.

Product-Specific Past 30-Day Flavor Type Among Past 30-Day Users

Past 30-day flavor types among past 30-day tobacco product users are shown (with a minimum of 50 unweighted users) across all age groups and products (Table S3) (see Table 2 for prevalence among past 30-day FTP users).

Menthol/mint.

In the past 30 days, adults 25+ used menthol/mint e-cigarettes at higher rates than youth and used menthol/mint any cigar at higher rates than YAs. Youth and YA past 30-day cigarette smokers had more prevalent use of menthol/mint cigarettes than adults 25+.

Fruit.

Compared with adults 25+, youth had a higher prevalence of past 30-day fruit flavor use for cigarettes, e-cigarettes, traditional cigars and any cigar.

Candy/sweet.

In the past 30 days, youth disproportionately used candy/sweet flavors in cigarettes, e-cigarettes, any cigar, and hookah, compared with adults 25+.

Concordance of Self-Reported and Expert-Coded Brand-level Flavor Use and Type

Table 3 shows weighted data representing the 10 most popular expert-coded primary flavor for cigarettes, traditional cigars, cigarillos, filtered cigars, pipe, hookah, smokeless tobacco, and snus by self-reported flavor type for all adult (ages 18+) established tobacco product users (reporting a usual/last brand).^d Brands were grouped according to the primary flavor indicated by expert brand coding. Self-reporting of brand flavor was largely consistent with expert coding of that brand as flavored or non-flavored. For cigarettes, over 90% of the top 10 brand expert-coded flavor type smokers identified their non-menthol brand as non-flavored; over 80% of comparable menthol smokers identified their brand as flavored. For smokeless, fewer non-flavored users reported their brand as non-flavored (over 50%), compared with menthol/mint (over 70%) or fruit flavor (over 90%) brand users reporting their brand as flavored. Almost all top 10 hookah brands had a primary fruit flavor; depending on brand, 41% to 82% of these hookah users reported their brand as flavored. For cigarillos and hookah, adults noted multiple flavor types of their own brand beyond what brand labeling/marketing would suggest as the primary flavor. For example, cigarillos and filtered cigars expert-coded as having a primary fruit flavor were most commonly self-reported by participants as being fruit-flavored, followed by candy/sweet flavor (see Table 3). Hookah brands expert-coded with a primary fruit flavor were also most commonly self-reported as being fruit-flavored, and then depending on brand, followed by candy/sweet (15.5–31.5%), mint/menthol (14.0–28.2%), some other flavor (7.3–14.8%) and alcohol (6.6–11.5%).

Table S4 shows unweighted data representing concordance by product of self-reported and expert panel-coded brand flavor use for adults 18+. For any flavor use versus non-flavored use, we found substantial concordance for cigarettes (κ = 0.87) and smokeless tobacco (κ = 0.62) and moderate concordance for traditional cigars (κ = 0.53), cigarillos (κ = 0.48), filtered cigars (κ = 0.54), pipe (κ = 0.48), and snus (κ = 0.49). We found low concordance for hookah (κ = 0.01), which the expert panel more often noted as flavored, after review of brand descriptors and marketing materials, than did respondents.

Discussion

This national study found that FTP use remains high among U.S. tobacco users across age groups and that many flavor types, not just sweet flavors,³⁷ appeal to young tobacco users. Individuals of all ages used fruit, menthol/mint, and candy/sweet flavors most commonly at first and past 30-day use. There was an inverse age gradient in use of clove/spice, fruit, chocolate, alcoholic drink, candy/sweet, and other flavor types with higher prevalence among younger users. Menthol/mint use overall (though not for cigarettes) showed a positive age gradient and was prevalent across age groups (>50% youth, YAs; 76%, adults 25+).

Like the National Adult Tobacco Survey, flavor type prevalence varied across tobacco products,⁸ with high prevalence of menthol/mint for past 30-day cigarette and smokeless tobacco (excluding snus pouches) use and high prevalence of fruit flavor use for past 30-day cigarillo, hookah, and e-cigarette use. These data are consistent with convenience store sales showing the highest market share for fruit-, sweet-, and wine-flavored cigars,¹² but diverge from flavored e-cigarette convenience store and other retailer sales showing the greatest market share for menthol/mint e-cigarettes.^{38, 39} However, these store data exclude vape shops, which offer a large number of e-cigarette flavors.^{40, 41} Although clove cigar sales increased after the flavored cigarette ban,⁴² first and past 30-day use of clove/spice FTPs across ages was low. We did not assess flavored kretek (a clove cigarette) though use is unlikely to affect these estimates. E-cigarettes and hookah showed the highest prevalence of flavor use, consistent with high flavor use for these products and cigars in the National Youth Tobacco Survey.⁴

Our results suggest that adults (18+) can accurately self-report flavored and non-flavored tobacco product use. However, for some products, like cigarillos and hookah, users reported more flavor types than are evident from primary flavor descriptors or marketing materials. This may be because products without characterizing flavor names may still contain high levels of flavor compounds.⁴³ Respondents may also perceive flavoring based on features other than flavor descriptors (e.g., colors, smell) which can affect product perceptions.^{44, 45} Analyses based solely on flavor name descriptors may incorrectly estimate the amount of flavoring in non-cigarette tobacco products in either direction. To better identify FTPs, analyses could compare user perceptions with information on brands/sub-brands, as in the PATH Study.

Study strengths include assessment of flavor types used by a large national sample of youth, YAs, and adults 25+, examining multiple measures of first, past 30-day, and brand use; comprehensive sub-brand assessment including pictures; and comparison of expert-coded and self-reported brand flavor measures not previously conducted. This study has several limitations. Perception and recall of first or past 30-day FTP use or type may affect these estimates. Youth may lack a usual brand and use multiple brands, giving them more opportunity to use multiple flavor types. This may inflate the observed age gradient; however, we also found an age gradient among younger versus older adults. We also note a decrease in youth FTP use from Wave 1,¹³ but do not examine longitudinal switching to non-FTP use. However, studies have found that more menthol cigarette smokers switched to non-menthol cigarettes than the reverse.^{46, 47} We also reported any new tobacco product use but not the very first product used. Existing FTP users may prefer FTPs when starting new tobacco products. However, because we report only new product use since Wave 1, this avoids a cohort effect of different tobacco products available to younger versus older users at first use. We additionally conducted concordance analyses only among adult established users since youth were not asked about brand flavor type. Adults with a preferred brand may differ from adults without one in FTP use or recall of flavor type, limiting generalizability. Next, we lack information on flavor additives which may affect perceptions of flavor type. Finally, this study utilizes 2014–15 data, and flavor use or type may have changed due to shifts in use patterns and the tobacco marketplace such as increasing youth e-cigarette use or the rise of newer e-cigarette pod products.^{48, 49}

Conclusions

Evaluating both consumer perceptions of flavor types and flavor descriptors may better identify FTP use patterns. Future research can examine trends in flavor type use across products and ages to determine how FTP use influences tobacco use trajectories over time. Information regarding specific flavor types used by product and age can also inform federal and state/locality regulations. For example, in March 2019, FDA published draft guidance for industry proposing modifications to compliance policies for flavored electronic nicotine delivery system (ENDS) products (other than tobacco, mint, and menthol flavor) and flavored cigars (other than tobacco flavor).⁵⁰ The proposal prioritizes enforcement resources for flavored ENDS and cigars that do not have proper marketing authorizations from FDA and are sold in ways that increase risk for youth access. Such approaches can help reduce youth tobacco use.

What this paper adds.

• Among new product and past 30-day users of flavored tobacco, menthol/mint, fruit, and candy/sweet flavors were commonly reported flavor types used by youth, YAs, and adults 25+.

• For overall flavored tobacco use in the past 30 days, youth and YAs had a significantly higher prevalence of use of each flavor type, compared with adults 25+, except menthol/mint, which had high prevalence among all ages but highest prevalence use among adults 25+, followed by young adults, and then youth users, though the reverse pattern was found for cigarettes. No flavor type was uniquely used by adults 25+ only.

• There was moderate to substantial concordance of brand- and self-reported FTP use across most tobacco products. However, particularly for hookah and cigarillo flavored brands, adults self-reported using a greater range of flavor types than is evident from primary flavor descriptors in brand marketing information.