RESEARCH REPORT

doi:10.1111/j.1360-0443.2010.03201.x

The relationships between menthol cigarette preference and state tobacco control policies on smoking behaviors of young adult smokers in the 200607 Tobacco Use Supplements to the Current Population Surveys (TUS CPS)
Karen Ahijevych & Jodi Ford
The Ohio State University, College of Nursing, Columbus, Ohio, USA

ABSTRACT Aim To examine relationships between the preference for menthol cigarettes and young adult smoking behaviors, including the extent to which state tobacco control policies moderate these relationships. Design Cross-sectional design using secondary data from the 200607 Tobacco Use Supplements to the Current Population Surveys (TUS CPS) surveys appended with 2006 state-policy data. Setting United States nationally representative survey. Participants A total of 2241 young adult daily smokers and 688 young adult non-daily smokers. Measurements The two dependent variables of smoking behaviors were smoking rst cigarette within 30 minutes of waking (TTF) and number of cigarettes smoked per day (cpd). Primary independent variables included menthol brand preference and state tobacco control policies (youth access laws, clean indoor air laws and cigarette excise taxes), adjusting for controls. Findings Among daily smokers, there were no signicant associations between menthol brand preference and TTF or cpd. However, lower educational attainment, not being in the labor force and the lack of home smoking rules were associated positively with shorter TTF, being white and the lack of home smoking rules were associated positively with cpd. Among daily smokers, state excise taxes were associated negatively with higher cpd. Among non-daily smokers, menthol brand preference was associated positively with shorter TTF, but associations did not vary with state tobacco control policies. Menthol brand preference was not associated signicantly with cpd, but male gender, unmarried status and the lack of home smoking rules were associated positively with greater cpd among non-daily smokers. Conclusions Young adult non-daily smokers who preferred menthol cigarettes were signicantly more dependent than those who preferred non-menthol cigarettes, as shown through the shorter TTF. Associations between menthol brand preference and smoking behaviors did not vary with state tobacco control policies. Keywords Mentholated cigarettes, nicotine dependence, young adult smokers, tobacco control policies.

Correspondence to: Karen Ahijevych, The Ohio State University, College of Nursing, 1585 Neil Avenue, Columbus, Ohio 43210, USA. E-mail: ahijevych.1@osu.edu Submitted 29 July 2010; initial review completed 12 March 2010; nal version accepted 6 August 2010

INTRODUCTION Young adults are at high risk for smoking due to increasing autonomy from parents, reaching legal age to purchase tobacco products, increasing use of other addictive substances and direct targeting by the tobacco industry to sustain tobacco use and recapture those who quit smoking [13]. According to national estimates from 200507, nearly 24% of young adults aged 1824 years reported currently smoking every day or on some days the highest smoking prevalence among all age groups
2010 The Authors, Addiction 2010 Society for the Study of Addiction

[4]. In addition, young adult smokers preference for menthol cigarettes increased from 34.1% to 40.8% between the years 200408 [5], which is concerning, because research suggests that menthol cigarettes may be a starter tobacco product and increase nicotine dependence [6]. Evidence of the role of menthol brand preference in shaping the smoking behaviors of young adults is limited, but research has found that adolescent menthol smokers have 45% higher odds of being above the median on nicotine dependence and are signicantly less likely to consider quitting smoking than teens who smoked
Addiction, 105 (Suppl. 1), 4654

Young adult smokers and menthol cigarettes

47

non-menthol cigarettes [6]. Research has found that people who quit smoking before the age of 35 have a life expectancy similar to those who never smoked [7]. Therefore, identifying variables related to nicotine dependence among young adult smokers may inform cessation strategies as recommended in the Institute of Medicine (IOM) report on ending the tobacco problem [8]. However, our understanding of the factors that prevent smoking or reduce nicotine dependence among young adults is limited due to a scarcity of research among this population [9], particularly among higherrisk groups who already smoke on a daily or occasional (non-daily) basis, as well as between menthol and nonmenthol users. Research among adolescents and young adults suggests that state tobacco control policies, particularly cigarette taxes, may be effective in reducing the prevalence of smoking and increasing tobacco cessation [10,11]. However, research also found that increased cigarette taxes were associated with a reduction in the number of cigarettes smoked among young adults [12,13], but the increased tax also was associated with young adults switching to cigarette brands with higher amounts of tar and nicotine [12]. Strong clean indoor air laws and increasing cigarette prices were associated with decreased cigarette consumption in a study among current smokers aged 1580 years [13], while another found that young adults were more likely to quit smoking when exposed to more restrictive smoking bans in the work-site and public places [11]. Youth access policies and enforcement aim to reduce sales of cigarettes to minors, but studies provide only limited support of their effectiveness [14]. However, adolescents obtain most of their cigarettes from non-retail sources, such as older peers and parents or through theft; thus the effectiveness of state policy controls may be attenuated [14]. Because young adulthood is a critical time for the establishment of nicotine dependence [15], efforts to reduce smoking among this vulnerable population are imperative. Further research on the role of state tobacco control policies and nicotine dependence among young adults is needed, particularly among specic population groups and types of users. Therefore, the purpose of this research was twofold: [1] to examine associations between menthol brand preference and the smoking behaviors of young adult daily and non-daily smokers and [2] to examine the extent to which the associations between menthol brand preference and smoking behaviors were moderated by state tobacco control policies. METHODS Study design and data sources A cross-sectional design was employed using secondary data from the Tobacco Use Supplements to the Current
2010 The Authors, Addiction 2010 Society for the Study of Addiction

Population Surveys (TUS CPS)May 2006, August 2006 and January 2007 surveys appended with 2006 state-policy data on youth access laws, clean indoor air laws and cigarette excise taxes. The TUS is sponsored by the National Cancer Institute (NCI) and the Centers for Disease Control and Prevention (CDC) and surveys US households on smoking and other tobacco use [16]. The TUS is a supplement to the core CPS that the US Census Bureau conducts monthly to examine employment status and socio-demographic factors of US households. For this study, data on state-level youth access and clean indoor air laws were derived from NCIs State Cancer Legislative Database Program (http://www.scld-nci.net) and the data on state-level cigarette excise taxes were obtained from the American Lung Association (ALA; State of Tobacco Control). Sample The 200607 TUS CPS surveys were administered to approximately 267 000 civilian, non-institutionalized people. The sampling frame for our study was young adults aged 1824 years who reported smoking daily (n = 2339) or non-daily (n = 795), which was dened as smoking 129 of the last 30 days. This study utilized only data from self-respondents, because certain items of interest related to tobacco use were not asked of proxies. Observations with missing data were excluded from analysis for a nal sample size of 2241 young adult daily smokers (missing n = 98) and 688 non-daily smokers (missing n = 107) who lived across the 50 states and Washington, DC. The sample for this study was stratied by daily smoking and non-daily smoking for theoretical and methodological reasons. First, because menthol is thought to be a starter product to greater nicotine dependence [6], we wanted to capture the unique associations along this continuum. Secondly, non-daily and daily smokers most probably smoke for different reasons, with variations in risk and protective characteristics. Thirdly, although the TUS asks daily and non-daily smokers identical smoking behavior questions, items for the latter group refer to the average only on the days that they smoke, which complicates measurement and estimation if analyzed simultaneously. Measures Dependent variables Smoking behaviors were operationalized as smoking within 30 minutes upon waking and the average number of cigarettes smoked per day. Time to rst cigarette was measured as a categorical variable, with individuals who smoked their rst cigarette within 30 minutes of waking categorized as yes. The distribution for the average
Addiction, 105 (Suppl. 1), 4654

48

Karen Ahijevych & Jodi Ford

number of cigarettes smoked was non-normal even after taking the log; thus we analyzed this as a continuous count variable.

Independent variables Menthol brand preference was measured as a categorical variable, with yes indicating that young adults preferred to smoke only menthol brand cigarettes. Individual-level control variables included key socio-demographic characteristics and household smoking rules. Race and ethnicity were categorized as Hispanic, non-Hispanic white, non-Hispanic black and non-Hispanic other, with the nal category including those participants who selfidentied as non-Hispanic Asian, American Indian and multi-racial, due to small sample sizes. Foreign birth was categorized as born in United States versus born outside United States. Socio-economic position was measured via three indicatorsemployment status, household income and educational attainment. Employment status was categorized as employed, unemployed and not in the labor force during the week before the survey. Individuals were classied by the TUS CPS as not in the labor force if they were retired, disabled or other status, which included students or those participants who were running a household. Income in the TUS CPS is an ordinal measure of annual family income that was accrued during the 12 months preceding the survey, including income obtained from employment, business, farm or rent, dividends, social security, pensions and interest. We categorized the income measure into four brackets (as in Table 1). Lower educational attainment was measured as those participants who were aged 18 years with less than 11th grade education or aged 1924 years without a high school diploma. Presence of smoking rules in the household were measured categorically as no smoking rules in the home, some smoking allowed and no smoking allowed at all. Additional controls included age, gender, marital status (categorical measure of married versus never married, divorced, widowed or separated), and for nondaily smokers a control variable for the number of days that they smoked per month. Four state-level variables were examined: 2006 youth access tobacco laws, 2006 clean indoor air laws, 2006 cigarette excise tax and 200607 smoking prevalence. Youth access laws were measured using a composite developed by Alciati and colleagues [17] to rate the extensiveness to which states legislate tobacco use among youth. The composite rates nine tobacco control measures, including minimum age for purchase, sealed packaging, clerk intervention for purchase, photo identication for purchase, ban on the sale of tobacco products through vending machines, ban of free tobacco samples or rebates, graduated penalties for retailers violating the
2010 The Authors, Addiction 2010 Society for the Study of Addiction

youth access laws, random inspections of retailers and state-wide enforcement of the laws. Six of the items are rated from 0 (no effective provision) to 4 (meets target) and three have the possibility to score 5 for exceeding the target goal for a possible total of 39 on the composite. The state clean indoor air laws also were measured using a composite developed by Chriqui and colleagues [18] that rates the extensiveness to which states restrict indoor tobacco use. The composite rates nine tobacco control measures, including smoking restrictions in seven site-specic areas and two provisions on enforcement. Scoring of items range from 0 (no effective provision) to 4 (meets target) and if the state exceeds the target on six of the nine items, then they are eligible to receive an additional 6 points for a maximum score of 42. Data on state excise taxes represent the tax added to the purchase of a pack of cigarettes in US dollars. The variable was logged for the purposes of this analysis to achieve a normal distribution. Because state excise taxes increased among several states between data collection points in the 200607 TUS, we disaggregated this measure to the individual level to adjust for these changes. All state-level policy variables were continuous measures. A state-level measure of the prevalence of young adults (aged 1829 years) who smoked daily or non-daily was included as a control [19].

Analysis Descriptive and multi-level analyses were conducted using SAS version 9.1 statistical software (SAS Institute, Cary, NC, USA) and HLM 6.07 (SSI, Inc., Lincolnwood, IL, USA). Multicollinearity was examined and no inuential correlation between variables was found. All analyses were unweighted as multi-level modeling software packages are unable to accommodate replicate weights [20]. Descriptive analyses were conducted to yield characteristics of participants. Multi-level analyses were conducted to examine associations between menthol brand preferences and the dependent variables measuring smoking behaviors, including cross-level interactions with state-level tobacco control policies. Multi-level random intercept models were analyzed using the Bernoulli distribution for the binary dependent variable (rst cigarette smoked within 30 minutes of waking) and a Poisson distribution corrected for overdispersion with a negative binomial distribution for the count dependent variable (average number of cigarettes smoked). [We chose this analytical technique because it allows for random effect modeling inherent to the nested nature of our data. However, this model will produce a prediction for the zero count but a zero is not included in the distribution of the dependent variable, which ranges from 1 to 60. We analyzed the data using a zero-truncated negative binomial regression model with
Addiction, 105 (Suppl. 1), 4654

Young adult smokers and menthol cigarettes

49

Table 1 Descriptive ndings of socio-demographic characteristics and menthol brand preference of young adult daily and non-daily smokers, including state tobacco control policiesa. Daily smokers n = 2241 % Dependent variables First cigarette smoked within 30 minutes of waking Average number of cigarettes smoked daily Independent variables Menthol brand preference Race Black Hispanic Other White Foreign birth Socio-economic position Household income Missing $024 999 $2549 999 $50 000+ Lower education Employment Unemployed Not in labor force Employed Age Male Married Lives with parent Number of days smoked per month Home smoking rules Smoking allowed Some smoking allowed No smoking allowed State youth access laws State clean air laws State cigarette excise tax $ State logged cigarette excise tax $ State prevalence of current smoking
a

Non-daily smokers n = 688 Mean (SD) % Mean (SD)

50.3 29.9 6.4 7.0 6.7 79.9 4.5

13.5 (7.4)

9.7 25.7 8.1 14.5 8.0 69.4 11.5

4.4 (4.2)

7.0 43.1 28.5 21.4 24.1 12.7 19.2 68.1 48.9 14.9 28.9

6.1 43.5 26.4 24.0 16.7 10.8 17.7 71.5 48.8 11.9 24.7 11.8 20.6 67.6 19.3 23.3 0.96 -0.27 24.6 (6.4) (12.2) (0.59) (0.74) (5.7) 19.4 24.3 0.99 -0.21 23.3 (6.3) (12.1) (0.57) (0.71) (5.5)

21.5 (1.9)

21.8 (1.8)

13.1 (7.1)

30.5 21.3 48.2

Unweighted analyses. SD: standard deviation.

robust standard errors and the ndings were consistent between the two techniques.]

RESULTS Table 1 presents descriptive characteristics of young adult daily and non-daily smokers. Among young adult daily smokers, half the sample smoked their rst cigarette within 30 minutes of waking and the mean number of cigarettes smoked was 13.5 (range 160). Approximately 30% of the young adult daily smokers reported that they preferred to smoke menthol cigarettes. The majority was non-Hispanic white (80%) and only 4.5% was foreign-born. In relation to socio 2010 The Authors, Addiction 2010 Society for the Study of Addiction

economic position, 68% were employed, but 43% reported a low household income and 24% had lower educational attainment. The sample was nearly 50% male and only 15% was married. Nearly half lived in households where smoking was not allowed anywhere, 21% lived in households where smoking was allowed only in certain places and 30% lived in households where smoking was allowed in all places. Among non-daily smokers, on the days that they smoked approximately 10% smoked their rst cigarette within 30 minutes upon waking and the mean number of cigarettes smoked was 4.4 (range 130). Nearly 26% of non-daily smokers preferred menthol cigarettes. Approximately 70% of the sample was non-Hispanic white and
Addiction, 105 (Suppl. 1), 4654

50

Karen Ahijevych & Jodi Ford

Table 2 Random effects models of menthol brand preference and smoking behaviors of young adult daily smokers; 200607 Tobacco Use Supplement to the Current Population Survey (n = 2241)a,b. First cigarette within 30 minutes of waking Binary logit model Coefcient (SE) Level I xed effects Menthol brand preference Race Black Hispanic Other White (reference) Foreign birth Socio-economic position Household income Missing $024 999 $2549 999 $50 000+ (reference) Lower education Employment Unemployed Not in labor force Employed (reference) Age Male Married Lives with parent Home smoking rules Smoking allowed Some smoking allowed No smoking allowed (reference) Level II xed effects State youth access laws State clean air laws State logged cigarette excise tax $ State prevalence of current smoking Random effects Intercept Average number of cigarettes smoked daily Poisson model Coefcient (SE)

0.151 -0.102 -0.49 -0.192 ref -0.771

(0.088) (0.171) (0.170)** (0.214) (0.289)**

0.022 -0.292 -0.185 -0.220 -0.188

(0.031) (0.059)*** (0.050)*** (0.057)*** ref (0.062)**

0.182 0.076 0.085 ref 0.557 0.202 0.447 ref 0.019 0.538 0.096 -0.044 0.752 0.250 ref -0.007 -0.002 0.020 0.013 -0.815

(0.225) (0.145) (0.144) (0.105)*** (0.167) (0.134)** (0.022) (0.092)*** (0.135) (0.122) (0.096)*** (0.108)*

0.056 -0.023 0.028 0.055 0.042 0.023 0.021 0.176 -0.011 0.042 0.199 0.074

(0.041) (0.032) (0.033) ref (0.028) (0.042) (0.034) ref (0.006)** (0.026)*** (0.030) (0.034) (0.029)*** (0.028)** ref (0.002) (0.001) (0.017)** (0.003) (0.039)***

(0.006) (0.003) (0.051) (0.007) (0.134)***

-0.001 0.001 -0.046 0.004 2.43

*P < 0.05; **P < 0.01; ***P < 0.001. aUnweighted analyses. bFinal models. Cross-level interactions non-signicant, thus not depicted in model. Available upon request. SE: standard error.

11% was foreign-born. Socio-economic composition, gender and marital status were similar to daily smokers. Nearly 68% of non-daily smokers lived in households where smoking was not allowed anywhere, 21% lived in households where smoking was allowed only in certain places and 12% lived in households where smoking was allowed in all places. Table 2 presents the ndings of the multivariate analyses for both of the smoking behaviors among the sample of young adult daily smokers. Specically, in relation to time to rst cigarette, we found no associations between menthol brand preference and smoking within 30 minutes upon waking. In addition, no signicant associations between the state tobacco control policies and timing to rst cigarette were found, nor were there any
2010 The Authors, Addiction 2010 Society for the Study of Addiction

signicant moderating effect of policies on the associations between menthol brand preference and cigarette timing. Among control variables, we found that Hispanic (-0.49, P < 0.01) or foreign-born young adults (-0.77; P < 0.01) were less likely to smoke their rst cigarette within 30 minutes of waking compared to those who were white or US-born, respectively. In addition, young adult daily smokers with lower educational attainment were more likely than their more educated peers to smoke their rst cigarette within 30 minutes of waking (0.557; P < 0.001), as were those who were not in the labor force compared to those who were employed (0.447; P < 0.01). In addition, males were more likely than females to smoke their rst cigarette within 30 minutes (0.538; P < 0.001). Household smoking rules were related
Addiction, 105 (Suppl. 1), 4654

Young adult smokers and menthol cigarettes

51

signicantly to the timing of rst cigarette smoked. Specically, compared to young adult daily smokers who lived in households where no smoking was allowed, young adult daily smokers who lived in households with no smoking rules were more likely to smoke their rst cigarette within 30 minutes of waking (0.752; P < 0.001), as were those who lived in households where smoking was allowed in some places (0.250; P < 0.05). In relation to the average number of cigarettes smoked among daily smokers, no signicant differences in the number of cigarettes smoked daily were found between menthol and non-menthol cigarette smokers. Youth access and clean air laws had no signicant effect on the average daily number of cigarettes smoked. However, a small signicant effect was found for the logged state excise taxes. Specically, if state tobacco taxes were increased by 100%, the average number of cigarettes smoked among daily smokers would decrease by 4.6%. None of the tobacco control policies moderated associations between menthol brand preference and number of cigarettes smoked. Among control variables, young adult daily smokers who were Hispanic (-0.185; P < 0.001), black (-0.292, P < 0.001) or categorized as other (-220; P < 0.001) smoked fewer cigarettes than those who were categorized as white. Similarly, young adults who were foreign-born smoked fewer cigarettes than those who were US-born (-0.188; P < 0.01). Males smoked more cigarettes per day than females (0.176; P < 0.001) and the number of cigarettes smoked increased as age increased, although the effect was small (0.02; P < 0.01). Lastly, household smoking rules were related signicantly to the number of cigarettes smoked. Specically, the number of cigarettes smoked per day was higher among young adults who lived in households without any smoking rules (0.199; P < 0.001) as well as among those who lived in households where smoking was allowed in some places (0.074; P < 0.01), compared to young adults who lived in households where no smoking was allowed. Table 3 presents the ndings of the multivariate analyses for both of the smoking behaviors among the sample of young adult non-daily smokers. In relation to timing of rst cigarette, menthol users were found to be signicantly more likely to smoke within 30 minutes of waking compared to those who smoked non-menthol cigarettes (0.709, P < 0.05). However, non-daily smokers who lived in states with more strict clean air laws were more likely to smoke within 30 minutes of waking compared to those who lived in states with less strict laws (0.031, P < 0.05). No signicant cross-level interactions were found with state tobacco control policies and associations between menthol brand preference and timing to rst cigarette. Among control variables, Hispanic participants were less likely than non-Hispanic whites to smoke
2010 The Authors, Addiction 2010 Society for the Study of Addiction

within 30 minutes of waking (-1.41, P < 0.05), as were males compared to females (0.727, P < 0.01). In relation to the average number of cigarettes smoked among non-daily smokers, no signicant differences in number of cigarettes smoked were found based on menthol brand preferences or state smoking policies. In addition, no signicant cross-level interactions were found with state tobacco control policies and associations between menthol brand preference and timing to rst cigarette. Among control variables, Hispanic participants smoked fewer cigarettes per day on the days that they smoked than white participants (-0.445, P < 0.001), while males smoked more cigarettes than females (0.234, P < 0.01). In relation to socio-economic status, young adults who were not in the labor force smoked more cigarettes than those who were employed (0.239, P < 0.01). In addition, young adult non-daily smokers who were married smoked fewer cigarettes than those who were not married (-0.299, P < 0.01). Household smoking rules were associated signicantly with smoking, with a greater number of cigarettes smoked among young adults who lived in households without any smoking rules (0.199; P < 0.05) as well as among those who lived in households where smoking was allowed in some places (0.225; P < 0.01), compared to young adults who lived in households where no smoking was allowed. Finally, nondaily smokers who smoked more days per month smoked more cigarettes per day on average than those who smoked fewer days per month (0.022, P < 0.001). DISCUSSION The purpose of this study was to examine associations between menthol brand preference and smoking behaviors among young adult daily and non-daily smokers, including the extent to which state tobacco control policies moderated these associations. We found that menthol cigarettes were associated signicantly with shorter time to rst cigarette in non-daily smokers, but not among daily smokers. According to previous research, menthol cigarette use is more common among younger and less experienced smokers and greater nicotine dependence occurs among adolescent menthol smokers [6]. The ndings among non-daily smokers are in accordance with those of a laboratory study of adolescent smokers in which those who smoked menthol cigarettes had a signicantly shorter time to rst cigarette upon waking compared to non-menthol smokers (45% versus 29% smoking within the rst 5 minutes of the day, respectively), although no signicant differences in the number of cigarettes per day were found between the two groups [21]. Others report similar ndings, as African American and white women smoking menthol cigarettes daily had a signicantly shorter time to rst cigarette compared to non-menthol
Addiction, 105 (Suppl. 1), 4654

52

Karen Ahijevych & Jodi Ford

Table 3 Random effects models of menthol brand preference and smoking behaviors of young adult non-daily smokers; 200607 Tobacco Use Supplement to the Current Population Survey (n = 688)a,b. First cigarette within 30 minutes of waking Binary logit model Coefcient (SE) Level I xed effects Menthol brand preference Race Black Hispanic Other White (reference) Foreign birth Socio-economic position Household income Missing $024 999 $2549 999 $50 000+ (reference) Lower education Employment Unemployed Not in labor force Employed (reference) Age Male Married Lives with parent Number of days smoked per month Home smoking rules Smoking allowed Some smoking allowed No smoking allowed (reference) Level II xed effects State youth access laws State clean air laws State logged cigarette excise tax $ State prevalence of current smoking Random effects Intercept Average number of cigarettes smoked daily Poisson model Coefcient (SE)

0.709 0.262 -1.41 -0.298 ref -0.245

(0.317)* (0.457) (0.574)* (0.567) (0.495)

-0.018 -0.022 -0.445 -.140 ref -0.054

(0.110) (0.165) (0.064)*** (0.119) (0.120)

0.545 0.697 0.244 ref 0.498 -0.485 0.326 ref -0.052 0.727 0.032 0.056 0.050 0.620 0.330 ref -0.025 0.031 -0.153 0.007 -3.63

(0.746) (0.383) (0.506) (0.317) (0.507) (0.347) (0.083) (0.257)** (0.553) (0.376) (0.027) (0.427) (0.359)

0.218 -0.106 0.100 ref 0.123 0.070 0.239 ref -0.002 0.234 -0.299 0.051 0.022 0.199 0.225 ref -0.004 -0.007 0.018 0.006 1.26

(0.151) (0.114) (0.117) (0.093) (0.137) (0.086)** (0.018) (0.074)** (0.091)** (0.086) (0.005)*** (0.097)* (0.081)**

(0.024) (0.015)* (0.257) (0.034) (0.422)***

(0.005) (0.004) (0.049) (0.006) (0.109)***

*P < 0.05; **P < 0.01; ***P < 0.001. aUnweighted analyses. bFinal models. Cross-level interactions non-signicant, thus not depicted in model. Available upon request. SE: standard error.

smokers of both races [22]. Potential explanations for different effects of menthol cigarettes and TTF between daily and non-daily smokers may be due to differences in the measurement of time to rst cigarette and laboratory settings versus survey design. Specically, we used a categorical measure of having smoked the rst cigarette within the rst 30 minutes of waking in accordance with Fagan et al. [23], while the two aforementioned studies utilized continuous measures. Secondly, in relation to menthol exposure, research has found brand variability in the amount of menthol contained in cigarettes [24]. Because menthol additive masks harshness and discomfort of inhaling smoke, differences in the amount of menthol in a cigarette in combination with variation in
2010 The Authors, Addiction 2010 Society for the Study of Addiction

smoking topography behaviors may yield variability in menthol exposure.This may not be captured adequately in the categorical measure of menthol versus non-menthol cigarette information. No signicant variations in smoking behaviors were found between menthol users and non-menthol users based on differences in state tobacco control policies, which suggests that policies may have similar effects on smoking behaviors between these two groups. State policies related to clean air, youth access and cigarette excise tax had few signicant associations with the smoking behaviors of both daily and non-daily young adult smokers in our study overall. However, among non-daily smokers, clean air laws were associated with shorter time
Addiction, 105 (Suppl. 1), 4654

Young adult smokers and menthol cigarettes

53

to rst cigarette. In this case, individuals may smoke earlier in the day in anticipation of being in clean air facilities. Cigarette tax was the only state policy that had an effect on cigarettes per day, albeit a very slight reduction. The limited effect of state tobacco control policies in our study warrants further examination. First, in relation to cigarette taxes, price-sensitive smokers may use high price avoidance strategies such as a low or untaxed venue, discount or generic cigarette brands and coupons, which dampen the health impact of higher cigarette prices [25]. Secondly, policy effectiveness may vary based on psychosocial factors, as in one national study nding that adolescents who had low self-control were largely unresponsive to cigarette price [26]. Consequently, further study on how individual locus of control inuences smoking behaviors and the effectiveness of additional state tobacco control policies are warranted. Thirdly, young adults may be more responsive to social control measures against smoking if they occur in their immediate environment. Specically, home smoking bans had a signicant impact on smoking behaviors among our sample of young adults, which is in accordance with previous research [27]. Thus, increased emphasis on home smoking bans by peers, parents or partner may be an effective strategy to modify smoking behavior in the 1824-year-old age group of daily and non-daily smokers. Lastly, further research examining other contextual factors should be considered. For example, social norms related to menthol brand preference, tobacco industry marketing of menthol cigarettes, and anti-smoking media, including menthol messages may have more of an inuence on the smoking behaviors of young adults. Several of our ndings related to our control variables and smoking behaviors of young adult daily and nondaily smokers warrant further discussion. First, socioeconomic disparities were found with lower educational attainment associated with shorter time to rst cigarette among daily smokers, while non-daily smokers who were not in the work force smoked more cigarettes on average when they smoked compared to their employed peers. These ndings are in accordance with previous research [9,15,2830] and highlight the need to direct smoking cessation interventions to socio-economically disadvantaged populations. Additionally, we found that minorities, particularly young adult Hispanic smokers and those of foreign birth, were less likely than their white peers to be more dependent on nicotine, consistent with previous research [31]. However, more detailed investigations of Hispanics are needed, as previous research has found that health behaviors and health outcomes vary across Hispanics based on their country of origin, socioeconomic opportunities and acculturation [32,33]. Our study had several limitations. First, the study is cross-sectional, which precludes causal inferences. Longi 2010 The Authors, Addiction 2010 Society for the Study of Addiction

tudinal analyses on associations between menthol brand preference and lag effects of tobacco policies on smoking behaviors may yield more robust ndings. Secondly, a proportion of the TUS CPS respondents were accessed using random digit dialing computer-assisted telephone surveys, which may under-represent low-income populations and those engaging in higher risk behaviors, including smoking [33]. In both our samples, approximately 50% of the young adults were interviewed in person using computer-assisted devices, 43% were interviewed via telephone and 7% were missing on the response to type of interview. Increasingly, young adults are more likely than any other age group to have cell phone service only, thus the use of land-line telephone surveys for this population should be used with caution.Thirdly, our measures of time to rst cigarette and menthol use were limited to those in the TUS CPS, which did not include measures of smoking topography, exhaled carbon monoxide, plasma nicotine or cotinine concentrations. Consequently, our ability to detect signicant associations may have been restricted, and future studies should consider more rened measures. Despite these limitations, this study enhances our understanding of the role of menthol brand preference in the smoking behaviors of young adult daily and non-daily smokers, including no signicant differences in smoking behaviors between menthol and non-menthol users based on state tobacco control policies. Future research that examines the role of menthol cigarettes as starterproducts to greater nicotine dependence are needed, including policy analyses of state tobacco control efforts targeting menthol use. Declarations of interest None. References
1. Arnett J. J. Emerging adulthood: a theory of development from the late teens through the twenties. Am Psychol 2000; 55: 46980. 2. Park M. J., Mulye T. P., Adams S. H., Brindis C. D., Irwin C. E. The health status of young adults in the United States. J Adolesc Health 2006; 39: 30517. 3. Ling P. M., Glantz S. A. Tobacco industry research on smoking cessation: recapturing young adults and other recent quitters. J Gen Intern Med 2004; 19: 419 26. 4. Schoenborn C. A., Adams P. F. Health behaviors of adults: United States, 20052007. National Center for Health Statistics. Vital Health Stat 2010; 10(245). Available at: http://www.cdc.gov/nchs/data/series/sr_10/sr10_245. pdf. Accessed 4 June 2010. (Archived at WebCite at http://www.webcitation.org/5tevN98vL on 21 October 2010). 5. Substance Abuse and Mental Health Services Administration, Ofce of Applied Studies. The NSDUH Report: Use of
Addiction, 105 (Suppl. 1), 4654

54

Karen Ahijevych & Jodi Ford

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

Menthol Cigarettes. Rockville, MD: Substance Abuse and Mental Health Services Administration, Ofce of Applied Studies; 2009. Hersey J. C., Ng S. W., Nonnemaker J. M., Mowery P., Thomas K. Y., Vilsaint M.-Y. et al. Are menthol cigarettes a starter product for youth. Nicotine Tob Res 2006; 8: 40313. Doll R., Peto R., Boreham J., Sutherland I. Mortality in relation to smoking: 50 years observations on male British doctors. BMJ 2004; 328: 151927. Institute of Medicine. Ending the Tobacco Problem: A Blueprint for the Nation. Washington, DC: The National Academies Press; 2007. Ling P. M., Neilands T. B., Glantz S. A. Young adult smoking behavior. A national survey. Am J Prev Med 2009; 36: 389 94. Tworek C., Yamaguchi R., Kloska D. D., Emery S., Barker D. C., Giovino G. A. et al. State-level tobacco control policies and youth smoking cessation measures. Health Policy 2010; 97: 13644. e-pub ahead of print. Tauras J. A. Public policy and smoking cessation among young adults in the United States. Health Policy 2004; 68: 32132. Evans W. N., Farrelly M. C. The compensating behavior of smokers: taxes, tar, and nicotine. Rand J Econ 1998; 29: 57895. Chaloupka F. J., Wechsler H. Price, tobacco control policies and smoking among young adults. J Health Econ 1997; 16: 35973. Dinno A., Glantz S. Tobacco control policies are egalitarian: a vulnerabilities perspective on clean indoor air laws, cigarette prices, and tobacco use disparities. Soc Sci Med 2009; 68: 143947. Green P., McCausland K. L., Xiao H., Duke J. C., Vallone D. M., Healton C. G. A closer loom at smoking among young adults: where tobacco control should focus its attention. Am J Public Health 2007; 97: 142733. US Department of Commerce, Census Bureau. National Cancer Institute and Centers for Disease Control and Prevention Co-sponsored Tobacco Use Supplement to the Current Population Survey (200607). 2008. http:// riskfactor.cancer.gov/studies/tus-cps/. Data les (AND/OR) technical documentation (technical documentationwebsite). Available at: http://www.census.gov/apsd/techdoc/ cps/cpsmayaug06.pdf and/or http://www.census.gov/ apsd/techdoc/cps/cpsjan07.pdf. Accessed 11 November 2009. (Archived at WebCite at http://www.webcitation. org/5tex2ZfvB and at http://www.webcitation.org/5tex 7W8bu on 21 October 2010). Alciati M. H., Frosh M., Green S. B., Brownson R. C., Rosher P. H., Hobatrt R. et al. State laws on youth access to tobacco in the United States: measuring their extensiveness with a new rating system. Tob Control 1998; 7: 34552. Chriqui J. F., Frosh M., Brownson R. C., Shelton D. M., Sciandra R. C., Gobert R. et al. Application of a rating system to state clean indoor air laws (USA). Tob Control 2002; 11: 2634. Giovino G. A., Chaloupka F. J., Hartman A. M. Cigarette Smoking Prevalence and Policies in 50 States: An Era of ChangeThe Robert Wood Johnson Foundation ImpacTeen Tobacco Chart Book. Buffalo, NY: University of Buffalo State University of New York; 2009. Pierre F., Sadi A. 2008. Implementing resampling methods

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

for design-based variance estimation in multilevel models: Using HLM6 and SAS together. 2008 Joint Statistical Meeting, Section on Survey Research Methods. p. 788795. Available at: http://www.amstat.org/Sections/Srms/ Proceedings/y2008/Files/300724.pdf. Accessed 11 November 2009. (Archived at WebCite at http://www. webcitation.org/5teyFJqB3 on 21 October 2010). Collins C. C., Moolchan E. T. Shorter time to rst cigarette of the day in menthol adolescent cigarette smokers. Addict Behav 2006; 31: 146064. Ahijevych K., Parsley L. A. Smoke constituent exposure and stage of change in black and white women cigarette smokers. Addict Behav 1999; 24: 11520. Fagan P., Augustson E., Backinger C. L., OConnell M. E., Vollinger R. E., Kaufman A. et al. Quite attempts and intention to quit cigarette smoking among young adults in the United States. Am J Public Health 2007; 97: 141220. Kreslake J. M., Wayne G. F., Alpert H. R., Koh H. K., Connolly G. N. Tobacco industry control of menthol in cigarettes and targeting adolescents and young adults. Am J Public Health 2008; 98: 168592. Hyland A., Bauer J. E., Li Q., Abrams S. M., Higbee C., Peppone L. et al. Higher cigarette prices inuence cigarette purchase patterns. Tob Control 2005; 14: 8692. Fletcher J. M., Deb P., Sindelar J. L. Tobacco use, taxation and self control in adolescence. NBER Working Paper Series, 2009. Available at: http://www.nber.org/papers/w15130 Accessed 12 November 2009. (Archived at WebCite at http://www.webcitation.org/5teuSLXu0 on 21 October 2010). Clark P. I., Schooley M. W., Pierce B., Schulman J., Hartman A. M., Schmitt C. L. Impact of home smoking rules on smoking patterns among adolescents and young adults. Preventing Chronic Disease [serial online], 2006. Available from http://www.cdc.gov/pcd/issues/2006/apr/05_0028. htm Accessed 12 November 2009. (Archived at WebCite at http://www.webcitation.org/5tex5yyBx on 21 October 2010). Hu M.-C., Davies M., Kandel D. B. Epidemiology and correlates of daily smoking and nicotine dependence among young adults in the United States. Am J Public Health 2006; 96: 299308. Yang S., Lynch J., Schulenberg J., Diez-Roux A. V., Raghunathan T. Emergence of socioeconomic inequalities in smoking and overweight and obesity in early adulthood: the national longitudinal study of adolescent health. Am J Public Health 2008; 98: 46877. Ofce of Applied Studies. Past month cigarette use among racial and ethnic groups. National Survey on Drug Use and Health. 2006. Available at: http://www.oas.samhsa.gov/ 2k6/raceCigs/raceCigs.htm (accessed 21 October 2010). Zambrana R. E., Carter-Pokras O. Role of acculturation research in advancing science and practice in reducing health care disparities among Latinos. Am J Public Health 2010; 100: 1823. Lara M., Gamboa C., Kahramanian M. I., Morales L. S., Bautista D. E. H. Acculturation and Latino health in the United States: a review of the literature and its sociopolitical context. Annu Rev Public Health 2005; 26: 36797. Blumberg S. J., Luke J. V. Coverage bias in traditional telephone surveys of low-income and young adults. Public Opin Q 2007; 71: 73449.

2010 The Authors, Addiction 2010 Society for the Study of Addiction

Addiction, 105 (Suppl. 1), 4654