The Case to Include Brand of Moist Snuff in Health Surveys

Abstract

Introduction:

Brand of smokeless tobacco was added to the most recent Tobacco Use Supplement to the Current Population Survey (TUS-CPS), but deleted from the Centers for Disease Control’s National Adult Tobacco Survey. The objective of this study was to assess the utility of brand in distinguishing users of moist snuff.

Methods:

The sample consisted of participants from the 2010–2011 TUS-CPS who reported having used one of 14 brands of moist snuff in the past month ( n = 2334). The brands were categorized into one of three types: snus, discount snuff, premium snuff. Multinomial logistic regression was employed for testing for associations between brand type and a series of demographic and tobacco use measures.

Results:

Females, metropolitan residents, current smokers, and moderate users of snuff had significantly greater odds of using snus relative to premium snuff in the adjusted model ( P < .001). Frequent users of snuff (eg, daily users), current smokers, young adults, participants with low household income, and those exhibiting dependence on nicotine had greater odds of using discount versus premium snuff. Separate analyses among current smokers ( n = 470) and former smokers ( n = 70) revealed positive associations between smoking cessation attempts and smokers’ switch to discount snuff.

Conclusions:

Differences among the three categories of snuff users are likely attributed to variations in marketing campaigns. The differences are sufficient to warrant inclusion of snuff brand in health surveys because brand type could serve as a proxy measure for snuff use and dependence.

Implications:

Inclusion of brand of moist snuff in health surveys will enable researchers to categorize snuff users by brand type. Findings from this study indicate that brand type, defined according to cost (ie, discount vs. premium brands) and type of preferred snuff (ie, snus vs. other moist snuff), can distinguish snuff users by various demographic and tobacco use measures. Consequently, categorization by brand type could be used as a proxy measure for studies whose surveys do not include detailed information on snuff use and behavior.

Introduction

Assessment of smokeless tobacco (SLT) in health surveys has played a more prominent role in the tobacco control literature in recent years. The increased coverage is attributed to several factors such as the growth in marketing ¹ and sales of moist snuff ² ; the role of SLT in poly-tobacco use ³ ; the debate over tobacco harm reduction ⁴ ; and the emergence of low-nitrosamine snuff (ie, snus) in US markets. ⁵ While some of the epidemiologic studies have differentiated the three common forms of SLT (chewing tobacco, snuff, snus), few have made the distinction between discount and premium snuff. ² This is a critical distinction because discount snuff is associated with daily use and is popular among adolescents surveyed in a national study. ⁶ Other measures of snuff, such as flavoring and tobacco leaf-cut, could conceivably be used as indicators of the experience of a snuff user, but are seldom included in health surveys.

The importance of distinguishing snuff users is highlighted by three factors. First, brands of discount snuff now account for a large proportion of US market share (eg, 42% in 2011), driven largely by the increase in sales of Grizzly. ² Second, among the brands advertised in consumer magazines, discount brands were reported to have a higher concentration of unionized nicotine compared to premium brands. ⁶ The difference in nicotine concentration among brand types is highly relevant to the question of whether snuff use helps or hinders smoking cessation among dual tobacco users. ⁷ The third factor, taxation of snuff by price or weight, differs by US state and could potentially impact preference for one form of snuff over another. ⁸ The current study is intended to address the various issues by investigating demographic and behavioral characteristics among users of discount snuff, premium snuff, and snus. The results could support the decision to either include or exclude SLT brand in health surveys.

Methods

Participants

The sample was comprised of individuals who participated in one of three cross-sectional surveys administered by the US Census Bureau in May 2010, August 2010, and January 2011. ⁹ The study participants were civilian, noninstitutionalized adults (18+) who were randomly selected by household through multistage sampling. The participants were given a supplemental survey on tobacco use behaviors, referred to as the Tobacco Use Supplement to the Current Population Survey (2010–2011 TUS-CPS). The sample in the current study was restricted to self-respondents who used a brand of SLT that was classified as either snus or snuff ( n = 2334). Users of chewing tobacco were excluded because the tobacco is declining in market share and is taxed differently from moist snuff. This study also included SLT users who participated in the TUS-CPS panel study ( n = 141). These participants were assessed for brand switching between May 2010 and May 2011, but not for other outcomes due to the smaller sample size of the panel study.

Measures

The dependent variable had three nominal categories (discount snuff, premium snuff, snus) corresponding to the brand type used most often in the month prior to the survey. The survey included a list of SLT brands from which seven discount brands, four premium brands and three snus brands were identified for analysis (refer to footnote in Table 1 for list of brands). The discount and premium brands were differentiated by a cutoff of $3.50/can of moist snuff (at the time the survey), a value approximated from published data. ^{2 , 10} The independent variables in the primary analysis were gender, age group, education, household income, metropolitan residence, state taxation, smoking status, frequency of snuff use, and time to first use of snuff upon waking. The latter was excluded from multivariate analyses because its assessment among exclusive snuff users greatly reduced the sample size. The variable state taxation was defined according to a participant’s residence in a state that taxed snuff/snus by price (ad valorem), weight with a minimum tax, or weight without a minimum tax. Some states added a minimum tax as a disincentive for purchasing low-weight snuff (eg, snus).

Table 1.

Odds of Using Discount Snuff and Snus Relative to Premium Snuff Among Participants of the 2010–2011 Tobacco Use Supplement to the Current Population Survey ( n = 2334)

Independent Var.	%	Unadjusted OR ( SE )		Adjusted OR ( SE )
		Discount	Snus	Discount	Snus
Gender
Female (ref)	3.2
Male	96.8	1.33 (0.20)	0.23 (0.04)***	1.30 (0.20)	0.34 (0.07)***
Age group
18–25 y (ref)	22.2
26–49 y	58.7	0.88 (0.07)	0.50 (0.06)***	0.94 (0.07)	0.79 (0.12)
≥50 y	19.1	0.74 (0.07)**	0.47 (0.07)***	0.73 (0.07)**	1.03 (0.20)
Education
<HS diploma (ref)	11.9
HS diploma	40.4	0.85 (0.08)	0.52 (0.09)***	0.84 (0.09)	0.73 (0.13)
≥Some college	47.7	0.73 (0.08)**	0.73 (0.10)*	0.80 (0.09)	1.05 (0.17)
Household income
<$50 000/y (ref) a	52.0
≥$50 000/y	48.0	0.65 (0.04)***	0.48 (0.05)***	0.65 (0.04)***	0.68 (0.08)**
Metropolitan CBSA b
Non-metropolitan (ref)	41.5
Metropolitan	58.5	0.87 (0.04)**	1.64 (0.18)***	0.90 (0.05)	1.75 (0.22)***
State taxation
Ad valorem (ref)	68.6
Weight (min) c	17.9	1.08 (0.07)	0.76 (0.12)	1.15 (0.08)	0.75 (0.15)
Weight (no min)	13.4	0.76 (0.06)**	1.01 (0.13)	0.75 (0.07)**	0.87 (0.13)
Smoking status
Never smk. (ref)	46.0
Former smoker	26.8	1.18 (0.08)*	0.80 (0.14)	1.24 (0.08)**	0.84 (0.15)
Current smoker	27.2	1.34 (0.10)***	7.62 (1.09)***	1.64 (0.13)***	4.82 (0.77)***
Moist snuff use
Non-daily (<12 ) d	24.0
Non-daily (≥12 )	15.8	1.48 (0.13)***	0.37 (0.05)***	1.72 (0.16)***	0.61 (0.09)**
Daily use	60.2	1.54 (0.12)***	0.14 (0.02)***	1.91 (0.16)***	0.32 (0.05)***
Time to first snuff e
> 30min (ref)	77.1
≤30 min	22.9	1.25 (0.09)**	0.35 (0.12)**	NA f	NA f

HS = high school; OR = odds ratio; SE = standard error. The seven discount brands are Grizzly, Husky, Kayak, Longhorn, Red Seal, Silver Creek, and Timber Wolf; the four premium brands are Copenhagen, Hawken, Kodiak and Skoal; and the three snus brands are Camel, Marlboro and Skoal.

^a Cutpoint of $50 000/y based on median household income.

^b Core-based statistical area.

^c Weight-based taxation with a minimum tax.

^d Use of snuff on less than 12 days in prior month.

^e Smaller sample size ( n = 1587) due to exclusion of other tobacco product (OTP) users (ie, cigarettes, cigars, pipes).

^f Not applicable due to inclusion of OTP users.

* P < .05; ** P < .01; *** P < .001.

An additional measure, past-year smoking cessation attempt, was examined in separate analyses among dual tobacco users (ie, current smokers; n = 470) and former smokers ( n = 70) who reported having smoked 1 year prior to the survey. The current and former smokers were asked if they had used SLT in a past-year quit attempt, corresponding to unsuccessful and successful quit attempts, respectively. The brand of SLT used in any past-year quit attempt was based on the brand used in the month prior to the survey. A quit attempt for the current smokers was defined as smoking cessation for at least 1 day in the year prior to the survey.

Statistical Analysis

Multinomial logistic regression was employed for modeling the independent variables as predictors of the nominal outcome for brand type. Using premium snuff as the reference category (Y=0), the two logit functions are represented by the equation: ln [P(Y= k ) / P(Y=0)] = β _k0 + β _k1 X ₁ + β _k2 X ₂ + …. + β _km X _m , where k = 1 (discount snuff) or 2 (snus), and X corresponds to the respective independent variable. Variance estimates and confidence intervals for regression coefficients were derived using replicate weights based on Fay’s balance repeated replication. ¹¹ Thus, for the complex survey data, tests of independence for categorical variables were based on the Rao and Scott second-order correction (ie, F -statistic with non-integer degrees of freedom). ¹² All analyses were conducted using the survey commands in STATA v12.1. ¹³

Results

The majority of SLT users reported having used premium snuff in the prior month (54.8%), followed by discount snuff (37.5%) and snus (7.7%). Among users of the premium brands, 49.4% and 42.6% of participants reported using Copenhagen and Skoal, respectively. The majority of discount brand users and snus users preferred Grizzly (60.1%) and Camel Snus (55.6%), respectively. A minority of the SLT users (23%) in the TUS-CPS panel study ( n = 141) reported switching brands between 2010 and 2011; even fewer (10%) switched to a different brand type (premium, discount, or snus).

While males composed a vast majority of SLT users (96.8%) in the cross-sectional study, they had a substantially lower odds compared to females of using snus relative to premium snuff (refer to Table 1 ). Snus users also differed from premium snuff users in household income, metropolitan residence, smoking status, frequency of SLT use, and dependence on nicotine (ie, time to first use). For example, current smokers had 4.8 times the odds of using snus relative to premium snuff. Also, participants who used SLT less frequently (<12 days in last month) and exhibited less nicotine dependence had greater odds of using snus.

Statistically significant differences were also observed between users of discount and premium snuff for several demographic and tobacco use measures. As indicated in the adjusted model in Table 1 , several groups had lower odds of using discount versus premium snuff. They were the older participants (≥50), participants with above median household income (≥ $50 000/y), and participants residing in a state that administered a weight-based tax on snuff (without a minimum). Unlike these groups, current and former smokers, frequent users of snuff, and participants with greater nicotine dependence had greater odds of using discount versus premium snuff.

The adjusted estimates for smoking status in Table 1 indicate that current versus never smokers had greater odds of using discount snuff or snus compared to premium snuff. In addition, former smokers were more likely to use discount snuff than premium snuff. But, these results do not identify which of the three brand types were used for smoking cessation by former smokers ( n = 70) and dual tobacco users ( n = 470; ie, those who used SLT and cigarettes). As indicated in Figure 1 , the discount snuff users accounted for a greater proportion of those who attempted to quit smoking by switching to SLT, relative to the discount snuff users from the other groups of dual users ( F = 14.4; P < .0001). A similar finding was observed for the former smokers who successfully quit cigarettes ( F = 18.4; P < .0001); the former smokers who used snus are illustrated in Figure 1 , but excluded from the test statistic due to their small sample size ( n = 2). The findings from former and current smokers indicate that quit attempts with SLT correlate with smokers’ switch to discount snuff.

Figure 1.

Past-year smoking cessation attempts among US snuff users participating in the 2010–2011 Tobacco Use Supplement to the Current Population Survey.

Discussion

Differences among the three categories of snuff users were observed for key demographic variables (eg, gender), frequency of snuff use/dependence, smoking status and smoking cessation attempts. The observed differences are consistent with the evolving marketing of SLT. ^{14 , 15} Use of snus was associated with the female sex, metropolitan residence, less household income, current smoking, and modest use of snuff. These characteristics correlate with marketing messages that were being conveyed to consumers close in time to the 2010–2011 TUS-CPS. Prior to the national launch of Camel Snus in early 2009, the product was primarily being marketed to cigarette smokers as evidenced by product placement in convenience stores. ⁵ Shortly after the launch, consumer magazines ran several advertisements that highlighted the benefits of using Camel Snus, such as the absence of secondhand smoke. Unlike the advertisements for traditional moist snuff, the advertisements for Camel Snus appeared in women’s magazines such as Glamour, Essence and Vogue. ¹⁶ Furthermore, among the print advertisements reviewed by Timberlake et al., ¹⁶ 52% touted the benefit of being spit-free and 37% appealed to the upscale urbanite, modern messages that likely account for the association between snus use and female sex in the current study. The association with lower household income may be temporary given the frequent promotions for Camel Snus, which were directed to cigarette smokers (eg, free sample with cigarette purchase) from 2008 through 2009. ¹⁶

The more frequent use and dependence on discount versus premium snuff raises the question of whether marketing appeal or nicotine content accounts for the differences in use of the two brand types. Analysis of cross-sectional data and absence of pertinent measures in the 2010–2011 TUS-CPS precluded such determination. Though, one could argue that the discount brands, which have high concentrations of unionized nicotine, ¹⁰ are being marketed to the more experienced users. Thus, the greater use of discount snuff would be a function of marketing, not a function of greater nicotine concentrations that contribute to more frequent use. This argument is supported by the discount brand Grizzly whose marketing conveys masculine themes rather than themes oriented to the novice user. For example, Grizzly used the slogan “fruit flavors are for lipgloss” to counter the growth of flavored sub-brands that are largely promoted by the premium brand Skoal. ² Conversely, one could argue that the discount brands are being marketed to the youth ⁶ whom subsequently become dependent on nicotine. Findings from the current study indicate that young adults have greater odds of using discount versus premium snuff. Irrespective of whether marketing factors or product ingredients account for differences in the use of snuff brands, inclusion of brand in health surveys can be used by tobacco control advocates for planning and health promotion. Another finding of interest to tobacco control advocates is the correlation between smoking cessation attempts and smokers’ switch to discount snuff. A proper analysis of whether the high nicotine concentration in discount snuff helps or hinders smoking cessation would need to be conducted with longitudinal data.

The primary limitation of this study was recent pricing changes that are diminishing the demarcation of discount and premium brands. ² Thus, the differences observed among snuff users may be less apparent over time. A related limitation was the predominance of Grizzly which constituted 60% of the discount brands used by participants of the 2010–2011 TUS-CPS. An increase in the pricing of Grizzly, such as the launch of “Premium Grizzly” in 2011, ² could further diminish the classification scheme used in this study. Though, it is worth noting that sub-brands representative of the new trend (eg, Skoal Xtra/“Premium Grizzly”) could be included and, hence, differentiated from the other brands in health surveys. A third limitation was the absence of tobacco leaf-cut, a measure that could be used in distinguishing types of snuff users. A fourth limitation was the assumption that former and current smokers, who switched to SLT for smoking cessation, used the same SLT brand that was reportedly used in the month prior to the survey.

The primary strength of this study was assessment of brand at a time when snuff in the United States was undergoing substantial changes in marketing and taxation. The inverse correlation between use of discount snuff and residency in a state that taxes snuff by weight (without a minimum tax) may reflect the recent change in tax policy. ⁸

Conclusions

Notable differences by brand type provide support for the inclusion of snuff brand in health surveys. Brand type could be used as a proxy measure for frequency of use and dependence on moist snuff, highlighting the CDC’s need for continued assessment of brand in its National Adult Tobacco Survey. Even if pricing changes affect classification of the discount and premium brands over time, inclusion of brand could be used for differentiating snuff users on the basis of other factors such as nicotine content.

Funding

This work was supported by the National Institute of Drug Abuse and the Family Smoking Prevention and Tobacco Control Act (R03DA027950). The content is solely the responsibility of the author and does not necessarily represent the official views of the NIH or the Food and Drug Administration.

Declaration of Interests

None declared.