Tobacco Direct Mail Marketing and Smoking Behaviors in a Cohort of Adolescents and Young Adults From the U.S. Upper Midwest: A Prospective Analysis

Abstract

Introduction:

We assessed the characteristics of adolescents and young adults who received tobacco direct mail materials and the association of receiving these materials with subsequent smoking behaviors.

Methods:

Adolescents from the upper Midwest region of the United States were sampled through clustered random sampling in 2000 and surveyed every 6 months. Participants (n = 3546) were asked at baseline (October 2006–March 2007) whether they had received direct mail materials from tobacco companies during the previous 6 months. Smoking behaviors were assessed 6 months later (April–September 2007). We assessed associations between demographics and receiving tobacco direct mail materials at baseline and the association of receiving these materials with smoking behaviors at follow-up, stratified by baseline smoking status.

Results:

Overall, 5.2% of nonsmokers and 23.9% of current smokers in our sample received tobacco direct mail materials during the past 6 months (2.6% and 17.1% among nonsmokers and smokers <18 years old, respectively; p < .05). Participants who were older and living with smokers were more likely to receive these materials (p < .05). Baseline nonsmokers who received these materials smoked more cigarettes during the previous 30 days at follow-up (p < .05); baseline current smokers who received these materials were less likely than those who did not to reduce the number of cigarettes smoked during the past 30 days at follow up (p < .05).

Conclusions:

Those younger than 18 years old (particularly smokers) in addition to older individuals in our regional sample were exposed to tobacco direct mail marketing. Exposure to this market strategy is associated with faster escalation of cigarette consumption among nonsmokers and lower likelihood of smoking reduction among smokers in this adolescent and young adult sample.

INTRODUCTION

Tobacco use remains an important public health issue in adolescents and young adults: 19.5% of 9th–12th graders reported current smoking in 2009, and 34.2% of young adults (ages 18–25) reported current smoking in 2010 (U.S. Department of Health and Human Services, 2012). Tobacco marketing has been shown to be associated with adolescent smoking and adult cigarette consumption (Biener & Siegel, 2000; National Cancer Institute, 2008). Direct mail marketing is one of the advertising strategies used by the tobacco companies. Since the 1998 Master Settlement Agreement restricted tobacco companies’ ability to advertise their products in major media channels (e.g., newspaper, magazines, billboards), direct mail has been the only media channel that showed an increase in cigarette advertising expenditure. According to the Federal Trade Commission, the expenditures on cigarette direct mail marketing increased from about US$37 million in 1997, to about US$133 million in 2001 (a 359% leap from 1997), and to about $51 million in 2011 (a 138% increase from 1997; these figures did not include costs for coupon redemption or gifts included in the mail) (Federal Trade Commission, 2013; Lewis, Yulis, Delnevo, & Hrywna, 2004). Receiving tobacco direct mail materials (e.g., coupons, advertisements, etc.) could have an impact on adolescents’ and young adults’ smoking behaviors. For example, direct mail coupons can induce experimentation with smoking among nonsmokers and shape repeated purchase behavior among current smokers (Rothschild & Gaidis, 1981). Given that 99% of smokers initiate smoking and become daily smokers before the age of 30 (U.S. Department of Health and Human Services, 2012), it is important to examine the influence of tobacco direct mail marketing in adolescents and young adults.

Little is known about prevalence and characteristics of adolescents and young adults who receive tobacco direct mail materials. One study found that 4.9% of 12- to 17-year-olds reported ever receiving direct mail materials from tobacco companies, and ever receiving these materials was associated with current smoking (Altman, Levine, Coeytaux, Slade, & Jaffe, 1996). The proportion of adolescents currently receiving tobacco direct mail materials is unknown. Another study showed that 15.2% of young adults (aged 18–24) reported receiving tobacco direct mail materials in the past 6 months but did not report characteristics of young adults who received these materials. (Lewis, Delnevo, & Slade, 2004). The prospective effect of receiving tobacco direct mail materials on subsequent smoking behaviors among adolescents and young adults has not been previously examined. We used longitudinal data from the Minnesota Adolescent Community Cohort (MACC) study to examine characteristics associated with receipt of tobacco direct mail materials and its influence on subsequent smoking behaviors among adolescent and young adult nonsmokers and smokers.

METHODS

Study Population

The MACC study is a population-based prospective cohort study designed to further the understanding of the transitional process from nonsmoking to smoking during adolescence and to assess the effect of state- and local-level tobacco prevention and control programs on adolescents and young adults in Minnesota. The design of the study has been detailed elsewhere (Forster, Chen, Perry, Oswald, & Willmorth, 2011). Briefly, participants were selected in 2000–2001 and 2001–2002 from 60 randomly selected Minnesota geopolitical units (GPUs) and 5 GPUs from four comparison states (North and South Dakota, Michigan, and Kansas). Clearwater Research, Inc. used modified random digit dialing and a combination of probability and quota sampling methods to achieve an even distribution of adolescents from ages 12 to 16.

Of the eligible households, 3,636 participants in Minnesota and 605 participants in comparison states were recruited between 2000 and 2001 (recruitment rates of 58.5% and 58.3%, respectively). An additional cohort of 585 twelve-year-olds in Minnesota from the 60 previously randomly selected Minnesota GPUs was sampled and recruited using the same random digit dialing method during 2001–2002 (a recruitment rate of 63.6%), resulting in an overall sample of 4,826. Participants were surveyed periodically through 2011 using computer-assisted telephone interviews. In this analysis, we included participants who completed round 13 data collection (baseline for this analysis, conducted between October 2006 and March 2007, n = 3,526; response rate = 73.1% of the original cohort; 81.9% of those eligible for round 13 survey). Of these participants, 3,126 also completed round 14 data collection (follow-up for this analysis, conducted between April and September 2007; retention rate = 88.6%). Participants who were lost to follow-up were more likely to be from racial/ethnic minority groups, have lower education, and be heavier smokers (p < .01). We included participants residing in other states as well as Minnesota in the analysis because they did not differ in the prevalence of receiving tobacco direct mail materials (p = .68). The University of Minnesota Institutional Review Board approved this study. Participants (and their guardians if less than age 18) provided consent before completing the interviews, and monetary incentives ($10 for those aged 17 or under, $15 for those aged 18 or older) were provided.

Measures

Participants were asked at baseline if they had received an advertisement, a gift, or a coupon in the mail from a tobacco company in the past 6 months (yes/no). We assessed participants’ age, gender, race/ethnicity, and education at baseline, and participants reported whether they lived with a smoker and whether they had friends who smoke. Participants were asked the number of days they smoked in the past 30 days and number of cigarettes smoked on the days they smoked in the past 30 days both at baseline and follow-up. Those who reported smoking at least 1 day in the past 30 days at baseline were classified as current smokers, and those who did not smoke in the past 30 days at baseline were classified as nonsmokers. Baseline current smokers were further dichotomized into whether they smoked >15 days in the past 30 days (yes/no). We derived the number of packs of cigarettes smoked in the past 30 days at baseline and follow-up by multiplying the number of days smoked in the past 30 days and number of cigarettes smoked on the days they smoked and divided the product by 20. Among baseline smokers, we subtracted the number of packs of cigarettes smoked in the past 30 days at baseline from that at follow-up. Participants with a negative difference were classified as having reduced the number of packs of cigarettes smoked in the past 30 days at follow-up.

Statistical Analysis

We assessed the characteristics associated with receiving tobacco direct mail materials at baseline by including age, gender, race/ethnicity, education, living with a smoker, having friends who smoke, and “smoked more than 15 days in the past 30 days” among baseline current smokers in logistic regression models, stratified by baseline smoking status (nonsmokers vs. current smokers). To investigate the influence of receiving tobacco direct mail materials on progression of smoking behavior among nonsmokers, we assessed the association between receiving tobacco direct mail at baseline and the number of packs of cigarettes smoked in the past 30 days at follow-up using linear regression models, adjusting for age, gender, race/ethnicity, education, living with a smoker, and having friends who smoke. Among baseline current smokers, we assessed the associations between receiving tobacco direct mail materials at baseline and reduced number of packs of cigarettes smoked in the past 30 days at follow-up using logistic regression models, adjusting for age, gender, race/ethnicity, education, living with a smoker, having friends who smoke and whether the participants smoked >15 days in the past 30 days at baseline. Clustering by GPU was controlled for through specifying GPU as a random effect. All analyses were performed in SAS® v. 9.2 (SAS Institute Inc., 2009), and significance level was set at p < .05.

RESULTS

At baseline for this analysis (2006–2007), among baseline nonsmokers, 5.2% (n = 130) reported receiving tobacco direct mail materials in the past 6 months (>0%; p < .001; Table 1). Those under the age of 21 were less likely than those 21 and older to report receiving tobacco direct mail materials in the past 6 months (p = .008). However, 2.6% of nonsmoking adolescents under age 18 and 4.7% of nonsmoking 18- to 20-year-olds reported receiving tobacco direct mail materials (>0%; p < .02). Nonsmoking males were more likely than females to have received tobacco direct mail materials (p = .004). Nonsmokers who lived with smokers (vs. not living with smokers) and had friends who smoke (vs. not having friends who smoke) were more likely to have received tobacco direct mail materials (p < .05).

Table 1.

Correlates of Receiving Direct Mail (DM) Marketing Materials From Tobacco Companies During the Past 6 Months at Baseline (2006–2007) Among Nonsmokers and Current Smokers^a

Determinants	Nonsmokers			Current smokers
	Overall n	% receiving DM	AOR (95% CI)	Overall n	% receiving DM	AOR (95% CI)
Total	2,511	5.2%	–	945	23.9%	–
Age, years
Under 18	387	2.6%	0.29 (0.13, 0.72)	82	17.1%	0.42 (0.20, 0.88)
18–20	1,330	4.7%	0.59 (0.39, 0.88)	522	21.3%	0.66 (0.47, 0.91)
21 or above	794	7.2%	Ref.	341	28.2%	Ref.
Gender
Male	1,199	6.5%	1.69 (1.17, 2.44)	483	22.6%	0.90 (0.66, 1.22)
Female	1,312	4.0%	Ref.	462	24.2%	Ref.
Race/ethnicity
Non-Hispanic White	2,222	5.1%	Ref.	827	23.7%	Ref.
Other	289	5.5%	1.17 (0.67, 2.04)	118	21.2%	0.78 (0.48, 1.26)
Education
Enrolled in a 4-year college	1,052	4.4%	Ref.	224	14.7%	Ref.
Enrolled in a 2-year/technical college	283	7.1%	1.55 (0.89, 2.69)	135	24.4%	1.91 (1.10, 3.32)
Graduated from high school	549	7.3%	1.33 (0.84, 2.09)	387	28.2%	2.08 (1.31, 3.29)
Less than high school	627	3.8%	1.50 (0.80, 2.80)	199	23.1%	2.21 (1.24, 3.92)
Living with a smoker
Yes	572	10.0%	2.51 (1.73, 3.64)	464	26.5%	1.39 (1.02, 1.90)
No	1,939	3.8%	Ref.	481	20.4%	Ref.
Having friends who smoke
Yes	1,157	7.5%	1.92 (1.30, 2.83)	879	23.8%	0.95 (0.48, 1.88)
No	1,354	3.2%	Ref.	66	18.2%	Ref.
Smoked >15 days in the past 30 days
Yes	–	–	–	535	26.9%	1.28 (0.91, 1.79)
No	–	–	–	410	18.8%	Ref.

Note. AOR = adjusted odds ratio; CI = confidence interval.

^aEstimates were adjusted for all variables in the model. All proportions are significantly greater than zero (p < .05). Bolded ORs are statistically significant (p < .05).

At baseline of this analysis (2006–2007), among baseline current smokers, 23.9% (n = 221) reported receiving direct mail materials from tobacco companies in the past 6 months (>0%; p < .001; Table 1). Although those under the age of 21 were less likely than those older than 21 to report receiving these materials (p < .001), 17.1% of smokers younger than 18 and 21.3% of 18- to 20-year-old smokers reported receiving direct mail materials from tobacco companies (>0%; p < .05). Current smokers who had lower education (vs. those enrolled in a 4-year college) and living with other smokers (vs. not living with other smokers) were more likely to have received these materials (p < .05)

Pertaining to smoking behaviors at follow-up, we found that baseline nonsmokers who received tobacco direct mail materials reported smoking more packs of cigarettes in the past 30 days than those who did not receive these materials (0.34 vs. 0.13 packs; adjusted regression coefficient = 0.21, 95% CI [0.01, 0.41]). Baseline current smokers who received tobacco direct mail materials at baseline were less likely than those who did not receive these materials to have reduced their number of packs of cigarettes smoked in the past 30 days at follow-up (35.5% vs. 43.9%; adjusted odds ratio = 0.69, 95% CI [0.48, 0.99]).

DISCUSSION

Few studies have examined population exposure to tobacco direct mail marketing. We found that, overall, about 1 out of 20 nonsmoking and 1 out of 4 current smoking adolescents and young adults in our regional sample reported receiving tobacco direct mail materials in the previous 6 months. Most concerning is that about 5% of adolescents (regardless of their smoking status) reported receiving these materials in the past 6 months. Although it is possible that the materials were sent to the adult family members in the household, the fact that these adolescents remembered seeing them indicated they were exposed to this marketing strategy. Tobacco companies claim that they limit their direct mail marketing to those 21 years old or above (Philip Morris USA, 2012). However, we found that many of those under 21 years reported receiving these materials: 4% of nonsmokers and 21% of current smokers. Although it is possible that these participants used older adults’ identities to register for tobacco direct mail marketing, our data suggested that the Master Settlement Agreement and tobacco companies’ self-regulations are not enough to protect adolescents and 18- to 20-year-olds from tobacco marketing. Overall, the proportion of young adult participants who receive tobacco direct mail materials in our sample is higher than a previous report (which was 15.2%) (Lewis, Delnevo, et al., 2004).

Male nonsmokers and less educated smokers were more likely to report receiving direct mail materials from the tobacco companies, suggesting that tobacco companies may be targeting these subpopulations. The association between education and receiving tobacco direct mail materials may partly explain the tobacco use disparity for adolescents and young adults of lower socioeconomic status (U.S. Department of Health and Human Services, 2012).

To our knowledge, this is the first prospective analysis of the influence of receiving tobacco direct mail materials on subsequent smoking behaviors. We found that having received tobacco direct mail materials at baseline was associated with higher cigarette consumption at follow-up among baseline adolescent and young adult nonsmokers and lower likelihood of smoking reduction at follow-up among baseline adolescent and young adult smokers. This can be partly explained by the fact that 87% of tobacco direct mail materials collected between July 2011 and June 2012 by the Association for Nonsmokers–Minnesota contains coupons for tobacco products (Brock & Moilanen, 2012). As posited by marketing theory (Rothschild & Gaidis, 1981), providing coupons can induce purchase among first-time users and consolidate repeated purchasing behavior among current users. If this is true, policy interventions to eliminate tobacco coupons may be effective in reducing the influence of tobacco direct mail marketing on adolescents and young adults.

One limitation of the study is that attrition may introduce bias in our estimates. Given those who were lost to follow-up were more likely to be smokers, which is associated with receiving tobacco direct mail materials, we may have underestimated the prevalence of receiving these materials. However, because we adjusted for variables associated with dropping out from the study in the regression models, our estimates from the regression models should not be severely biased. Our regional and predominantly non-Hispanic White sample (reflective of the upper Midwest population) also limits the generalizability of our findings to regions of greater racial/ethnic diversity. Despite these limitations, our findings suggest that the Master Settlement Agreement and self-regulation by the tobacco industry have not been effective in protecting adolescents and young adults from tobacco direct mail marketing. Better surveillance of this marketing strategy is necessary to confirm our findings and to determine if policy interventions are warranted.

FUNDING

This work was supported by the National Cancer Institute at the National Institutes of Health (R01 CA86191 to JF, Principal Investigator).

DECLARATION OF INTERESTS

None declared.