While tobacco use in Australia

Transcription

1 Density of tobacco retail outlets near schools and smoking behaviour among secondary school students Abstract Objective: To investigate whether the density of tobacco retail outlets near schools in Victoria, Australia, is associated with adolescent smoking behaviour. Methods: Cross-sectional survey data of 2,044 secondary school students aged years was combined with tobacco outlet audit data. Associations between students self-reported tobacco use and the density of tobacco outlets near schools was examined using multilevel logistic and negative binomial regression models, with cigarette price at local milk bars and key socio-demographic and school-related variables included as covariates. Results: Increased tobacco retail outlet density was associated with a significant increase in the number of cigarettes smoked in the previous seven days among students who smoked in the past month (IRR=1.13; 95%CI ), but not the odds of smoking in the past month in the larger sample (OR=1.06; 95%CI ), after controlling for local mean price of cigarettes and socio-demographic and school-related variables. Conclusions and implications: This study suggests there is a positive association between tobacco retail outlet density and cigarette consumption among adolescent smokers, but not smoking prevalence, in the Australian context. There is value in considering policy measures that restrict the supply of tobacco retail outlets in school neighbourhoods as a means of reducing youth cigarette consumption. Key words: smoking, adolescents, retail marketing Aust NZ J Public Health. 2013; 37:574-8 doi: / Maree Scully, Molly McCarthy, Meghan Zacher, Charles Warne, Melanie Wakefield, Victoria White Centre for Behavioural Research in Cancer, Cancer Council Victoria While tobacco use in Australia has declined dramatically over the past 20 years in association with the introduction of comprehensive tobacco control policies, smoking remains the largest modifiable risk factor for death and disability in Australia. 1 In order to reduce the burden of smoking-related morbidity and mortality, research has sought to determine effective strategies for reducing uptake of smoking among young people. A number of demand reduction strategies have proven effective in reducing smoking uptake, including increasing the cost of tobacco through increases in tobacco tax excise, 2-4 and ensuring adequate exposure to anti-smoking mass media campaigns. 5,6 The enforcement of tobacco sales laws for minors is one of the few supply reduction strategies wielded in tobacco control and, while this policy is widely implemented, research has shown limited effects on youth smoking rates. 7 Supply restriction strategies are commonly used to reduce alcohol and drug consumption, and in the case of alcohol are supported by an evidence base which has established an association between the density of alcohol retail outlets and rates of alcohol consumption and experience of alcoholrelated harms among adults and young people in the surrounding community. 8,9 There is emerging evidence in tobacco control research of a similar association between the density of tobacco retail outlets and smoking rates, particularly in relation to the school neighbourhood and young people. A recent study by McCarthy et al. found that increased density of tobacco retail outlets within 1.6 km (1 mile) of Californian high schools was associated with a greater likelihood of experimental smoking (i.e. at least one cigarette in the past 30 days and less than 100 cigarettes ever smoked) among the local students. 10 Another Californian study found higher smoking prevalence at high schools in neighbourhoods with greater tobacco retail outlet density. 11 Research conducted in Ontario, Canada, found a positive association between the number of tobacco retailers surrounding a school and smoking susceptibility among students, 12 with a second study finding that secondary school students from areas of higher tobacco outlet density were more likely to report buying their own cigarettes. 13 Although international studies have found a positive association between tobacco retail outlet density in school neighbourhoods and adolescent smoking behaviours, it remains unclear what features of outlet density may be driving this association. None of the previous studies on the impact of tobacco retail outlet density noted above have accounted for the impact of local cigarette prices. With evidence of greater price sensitivity in young people, 2-4 it is possible that the pathway of influence may be through lower cigarette prices associated with increased local price competition, thereby leading to increased levels of consumption. 14 A further possible economic influence is the availability of tobacco, which is likely to increase Submitted: April 2013 Revision requested: June 2013 Accepted: August 2013 Correspondence to: Professor Melanie Wakefield, Centre for Behavioural Research in Cancer, Cancer Council Victoria, 1 Rathdowne Street, Carlton, Victoria 3053; Melanie.Wakefield@cancervic.org.au 574 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 2013 vol. 37 no. 6

2 Tobacco retail outlet density and smoking among students consumption by reducing the net costs (travel distance and related search costs) associated with obtaining the product. 14,15 Additionally, with evidence that exposure to tobacco point of sale (POS) in retail outlets can increase the perception that cigarettes are more easily accessible, 16 it may be that repeated exposure to tobacco retail outlets increases the likelihood of smoking by increasing the perception that acquiring cigarettes is easy. 17 The current study, conducted in 2008, sought to determine the relationship between the density of tobacco retail outlets near schools and adolescent smoking in Victoria, Australia, where a range of restrictions applied to tobacco retail POS displays, including a ban on tobacco advertising, and the restriction of the display of tobacco packs to one pack face per brand variant. Additionally, a number of factors found to be associated with smoking behaviour in adolescents were included in the model, such as socio-demographic variables and the mean local price of cigarettes, to allow delineation of the features of outlet density that might underlie an association with adolescent smoking. Method Design Cross-sectional survey data of secondary school students use of tobacco were combined with tobacco outlet audit data to examine the association between the density of tobacco outlets near schools and cigarette price at local milk bars (also known as delicatessens in some Australian states, i.e. smaller-sized general store outlets, typically family owned and run) and adolescent smoking behaviours. Adolescent smoking behaviour Data on adolescent smoking prevalence and cigarette consumption were obtained from the Victorian component of the 2008 Australian Secondary Students Alcohol and Drug (ASSAD) survey, a triennial national survey examining substance use among a representative sample of secondary school students. A total of 4,398 students in Years 7 to 12 from 67 randomly selected Victorian schools were surveyed between June and December Cancer Council Victoria s Human Research Ethics Committee approved the study and informed consent was obtained for all participants. This paper reports on data for 2,044 students aged 12 to 17 years from 35 schools in Victorian metropolitan areas (Melbourne and Geelong), for which tobacco outlet density and cigarette price information is available. The student participation rate for these 35 schools was 78%. The main reasons for students not participating in the survey were student absenteeism on the day of the survey and parental consent for participation not being returned to the school in time. Tobacco use was assessed by students indicating if they had smoked cigarettes in the last four weeks (defined as past month smoking) and the number of cigarettes smoked, if any, on each of the seven days preceding the survey. Individual-level control variables We used several additional pieces of information obtained through this survey as control variables, including age and sex. Students also recorded the amount of money they have available to spend on themselves (e.g. from pocket money, part-time job) during a normal week, how easy they think it is to buy cigarettes, and the smoking status of their parents/carers. Tobacco retail outlet density Information on the number of tobacco retail outlets within a 500 metre radius of the centre of each school was collected by pairs of trained field workers within two weeks of the students being surveyed. While larger buffer zones have been used to measure outlet density in previous studies, a 500 metre radius was chosen due to practical constraints. For the purposes of this study, tobacco retail outlets included milk bars, convenience stores, petrol stations and supermarkets. This definition encompassed the three most common retail outlets from which Victorian secondary school students who had smoked in the past seven days had purchased cigarettes (46% milk bars; 14% supermarkets; 13% petrol stations). 18 Cigarette price After counting nearby tobacco retail outlets, field workers identified the two closest milk bars to the school and returned to these locations to request permission from the store owners/ attendants to record information about their POS tobacco display. Where retailers declined to participate, field workers endeavoured to locate replacement milk bars. In order to maximise the number of schools with two milk bars audited, a larger inclusion zone (1 km radius of school) was employed. However, milk bars outside the 500 metre school radius were not included in the density measure. Of the 35 schools, two milk bars were audited for 26 schools, and one milk bar was audited for each of the remaining nine schools. At all audited outlets, information was collected on the price of a pack of 20 cigarettes from the two brands most commonly smoked by Victorian secondary students (Winfield and Peter Jackson). 18 Prices are in Australian dollars. Area-level deprivation A measure of neighbourhood socio-economic status (SES) was determined by matching school postcodes to index values assigned by the Australian Bureau of Statistics Socioeconomic Indexes for Areas (SEIFA) Index of Relative Socioeconomic Disadvantage. 19 This index ranks geographic areas (postcodes) on a continuum from high to low disadvantage based on area characteristics such as education and income. Schools ranked in the bottom third of the index were categorised as low SES, those in the middle third of the index as mid SES, and those in the upper third as high SES. Statistical analysis Tobacco outlet density was analysed as a continuous variable in one-outlet increments. To analyse cigarette price, we first computed the mean price of Peter Jackson 20s and Winfield 20s at the milk bar/s audited for each school. We then centred the data at the mean and rescaled so that a unit change corresponds to a $0.10 increase in regression models vol. 37 no. 6 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 575

3 Scully et al. As several variables (tobacco outlet density, cigarette price and SES) were measured at the school level rather than at the individual level, we used multilevel models (also known as hierarchical or mixed models) including a random intercept (also called random effect) for school to adjust standard errors for correlations both within and between schools. Multilevel logistic regression measured the effect of tobacco outlet density and price of cigarettes on the odds of past month smoking. Multilevel negative binomial regression was employed to model the number of cigarettes smoked in the previous seven days among smokers, as the count was skewed (Mean=18.7, SD=30.7, Median=5). Both models controlled for age, sex, SES, spending money, perceived ease of buying cigarettes, parental smoking status and school type. All analyses were conducted using Stata SE Results Descriptive data on the characteristics of the full student sample (n=2,044) and those smoking in the past month (n=269) are shown in Table 1. Among all students surveyed, the greatest proportion attended schools in high SES areas (53%), and 31% had more than $40 weekly spending money available. About one-third (34%) had at least one parent who smokes, and around one in four students (23%) believed that buying cigarettes was easy. Past month smoking prevalence was highest among: older students; those who went to school in less disadvantaged areas; those who had more weekly spending money; those who perceived buying cigarettes to be easy; and those who had at least one parent who smoked. The mean price of Peter Jackson 20s and Winfield 20s across schools was $9.43 (SD: $0.18; min: $9.05; max: $9.90) and the mean number of tobacco outlets within 500 m of a school was 2.37 (SD: 1.65; min: 0; max: 7). Among past month smokers, the mean number of cigarettes smoked in the previous seven days was 18.7 (SD: 30.7; min: 0; max: 189; median: 5), with 25% reporting not to have consumed any cigarettes in this period. Table 2 shows the results of fully adjusted multilevel logistic and negative binomial regression models predicting past month smoking and the number of cigarettes smoked in the previous seven days by smokers. After controlling for all covariates, neither tobacco outlet density (Odds Ratio [OR]=1.06; 95% confidence interval [CI] ); or cigarette price (OR=0.95, 95% CI ) was associated with the odds of past month smoking. When cigarette price was removed from the model, tobacco outlet density remained non-significant (OR=1.06, 95% CI ) and, similarly, cigarette price was not significant when tobacco outlet density was removed from the model (OR=0.94, 95% CI ). The number of cigarettes smoked in the previous seven days among past month smokers was not significantly associated with cigarette price (Incident Rate Ratio [IRR]=1.03, 95% CI ). However, it was predicted by tobacco outlet density, such that a unit increase in the number of tobacco outlets within 500 m of a school corresponded to an approximate 13% increase in the number of cigarettes smoked in the previous seven days (IRR=1.13, 95% CI ). Tobacco outlet density remained significant when price was removed from the model (IRR=1.14, 95% CI ), suggesting that the effect of density on cigarette consumption among adolescents is independent of prices at nearby tobacco retail outlets. In addition, the effect of price on past week cigarette consumption was not significant when outlet density was removed from the model (IRR=1.02, 95% CI ). Discussion The findings of the current study indicate that the density of tobacco retail outlets near secondary schools in metropolitan areas of Victoria is associated with past week cigarette consumption among adolescent smokers attending these schools after controlling for key socio-demographic variables and other known predictors. However, tobacco retail outlet density near schools was not associated with past month smoking. To our knowledge, this study is the first to examine the relationship between tobacco retail outlet density and adolescent smoking behaviour in the Australian context. Our results are in contrast to Table 1: Sample characteristics. Sample (n) Sample (%) Smoked last month (n) Smoked last month (% within group) Total 2, Age group years years years Sex Male Female 860 1, SES of school Low SES Mid SES High SES , Weekly spending money ($) None $1-$20 $21-$40 $41-$60 $61-$80 Over $ Perceived ease of buying cigarettes Not easy Easy 1, Parental smoking No parent smokes At least one parent smokes 1, SES = Socio-economic status. 576 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 2013 vol. 37 no. 6

4 Tobacco retail outlet density and smoking among students prior studies in the United States that have observed significant associations between the density of tobacco retail outlets near schools and past month smoking prevalence, 11 but not cigarette consumption, 10,11 among students. Some of this discrepancy might relate to differences in how individual-level and environmental influences were accounted for, with the current study controlling for local cigarette price and known predictors of smoking uptake, including available spending money, 20 parental smoking, 21 and perceived ease of access to cigarettes. 10,17 However, even when controlling only for age, sex, SES and school type, density was not significantly related to past month smoking prevalence (OR=1.04, 95% CI ), and was significantly and positively associated with past week consumption among smokers (IRR=1.14, 95% CI ). It could be that differences between Australian and American policy and environmental influences, which we cannot directly control for, are partially responsible for discrepancies between the current and previous studies. For example, at the time of data collection in Victoria and unlike the US tobacco advertising at retail outlets was banned, and graphic health warnings were present on all cigarette packets. These policy differences make the results of these studies difficult to compare directly. Further study of the relationship between outlet density and adolescent smoking in the Australian context is needed, particularly in light of additional Australian policy changes since 2008, including bans on the retail display of tobacco products and the introduction of plain cigarette packaging and larger graphic health warnings. The relationships between tobacco outlet density and past month smoking and past week consumption were independent of cigarette prices at nearby milk bars. This contradicts the suggestion that density is a proxy measure of increased price competition (and thereby lower cigarette prices). However, it should be noted that the variation in mean price of local cigarettes was relatively small, that price data were only collected once at each store, and that prices of only two brands were assessed, which may have reduced the ability to find an association between smoking behaviour and price. Thus, before discounting this hypothesis, it would be beneficial for future research in this area to use more comprehensive audit data on local cigarette prices. Lower net tobacco procurement costs (i.e. travel and other associated costs) are facilitated through increased tobacco retail Table 2: Results from multilevel logistic and negative binomial regression models examining predictors of past month smoking and number of cigarettes smoked in past week, including a random intercept for school. Logistic: Past month smoking Negative binomial: Number of cigarettes smoked in past week a n=2044 n=269 Main predictor variables Adjusted OR (95% CI) p value Adjusted IRR (95% CI) p value Number of tobacco outlets within 500m of school (1 outlet increase) 1.06 (0.90, 1.24) (1.02, 1.26) Cigarette price ($0.10 increase in average pack price) b 0.95 (0.83, 1.08) (0.95, 1.12) Covariates Age years years years 2.30 (1.38, 3.84) 2.95 (1.70, 5.13) 1.44 (0.82, 2.55) 1.41 (0.81, 2.48) Sex Male Female 1.47 (1.05, 2.06) 1.02 (0.77, 1.35) SES of school Low SES Mid SES High SES 1.64 (0.73, 3.68) 1.76 (, 3.09) 0.97 (0.59, 1.59) 1.24 (0.87, 1.76) Weekly spending money ($20 increase) 1.28 (1.16, 1.40) 1.09 (0.99, 1.20) Parental smoking No parent smokes At least one parent smokes 1.86 (1.37, 2.53) 1.41 (1.06, 1.88) Perceived ease of buying cigarettes Not easy Easy 2.81 (2.10, 3.77) 1.24 (0.95, 1.63) Notes: Both models adjusted for school type. OR = odds ratio; IRR = incident rate ratio; SES = socio-economic status. a Model includes students who smoked in the past month. b Average price of Peter Jackson 20s and Winfield 20s at audited milk bar(s) for each school, centred at the mean and analysed per $0.10 increase vol. 37 no. 6 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 577

5 Scully et al. outlet density, and this is a possible explanatory factor underlying the association between density and past week cigarette consumption. A number of previous studies have found a relationship between tobacco retail outlet density and neighbourhood characteristics associated with smoking, such as socioeconomic status and the ethnicity of the local population, suggesting that tobacco availability is likely to play a role in increased tobacco consumption. 14,15,22 Our findings further support the influence of tobacco availability on consumption. These findings provide some support for restricting the supply of tobacco retail outlets near schools as a means of reducing cigarette consumption levels among adolescent smokers. Licensing of retail tobacco outlets is a policy measure that can support restrictions on density, and is already legislated in some Australian jurisdictions (NSW, SA, WA and ACT). 23 However, these licences do not place restrictions on the density of outlets in any given catchment, or in areas where children and young people are most likely to be exposed. The restriction of tobacco retail outlet density through licensing has recently been recommended by the U.S. Institute of Medicine as a means to reduce tobacco consumption. 24 There are some limitations to this study that should be noted. (1) Only a relatively small number of schools were included in the sample; however, these schools were randomly selected and should be representative of metropolitan areas in Victoria, the second most populous state of Australia. (2) Adolescent smoking behaviour was based on self-reported data, which may be subject to social desirability and recall bias. (3) The measure of mean cigarette price was based on the price at the nearest milk bars of the two cigarette brands most commonly smoked by young people, and averaging these prices may have removed some of the price variation which combined with the fact that price data were only collected once at each store may have obscured its relationship with students smoking behaviours. (4) The measure of density did not include all tobacco retail outlets, instead only capturing the types of outlets at which young people are most likely to access cigarettes, 18 and the 500 metre buffer zone was smaller than those applied in previous studies conducted in the US and Canada (5) It was not possible to examine the impact of tobacco retail outlet density on experimental smoking separately to the influence it may have on more regular smoking as per McCarthy et al, 10 due to our comparatively smaller sample. (6) The use of cross-sectional data limits the ability to make causal inferences about the relationship between density and past week smoking consumption. This study adds to a growing body of international evidence demonstrating a positive association between tobacco retail outlet density in school neighbourhoods and smoking behaviours of young people. Given the significant health burden tobacco use imposes in Australia, there is value in considering policy measures that restrict the number of tobacco retail outlets near schools in an effort to reduce youth cigarette consumption. Licensing of retail tobacco outlets provides a policy platform through which density may be effectively restricted, and should be considered across all Australian jurisdictions. Acknowledgements The Victorian Department of Health contributed to the funding of the student survey. References 1. Australian Institute of Health and Welfare. Australia s Health Catalogue No.: AUS 99. Canberra (AUST): AIHW; Carpenter C, Cook PJ. Cigarette taxes and youth smoking: new evidence from national, state, and local Youth Risk Behavior Surveys. J Health Econ. 2008;27: Thomas S, Fayter D, Misso K, Ogilvie D, Petticrew M, Sowden A, et al. Population tobacco control interventions and their effects on social inequalities in smoking: systematic review. Tob Control. 2008;17: Franz GA. Price effects on the smoking behaviour of adult age groups. Public Health. 2008;122: National Cancer Institute. The Role of the Media in Promoting and Reducing Tobacco Use. Tobacco Control Monograph No.: 19. Bethesda (MD): US Department of Health and Human Services; Wakefield MA, Loken B, Hornik RC. Use of mass media campaigns to change health behaviour. Lancet. 2010;376: Fichtenberg CM, Glantz SA. Youth access interventions do not affect youth smoking. Pediatrics. 2002;109: Campbell CA, Hahn RA, Elder R, Brewer R, Chattopadhyay S, Fielding J, et al. The effectiveness of limiting alcohol outlet density as a means of reducing excessive alcohol consumption and alcohol-related harms. Am J Prev Med. 2009;37: Chen MJ, Grube JW, Gruenewald PJ. Community alcohol outlet density and underage drinking. Addiction. 2010;105: McCarthy WJ, Mistry R, Lu Y, Patel M, Zheng H, Dietsch B. Density of tobacco retailers near schools: effects on tobacco use among students. Am J Public Health. 2009;99: Henriksen L, Feighery EC, Schleicher NC, Cowling DW, Kline RS, Fortmann SP. Is adolescent smoking related to the density and proximity of tobacco outlets and retail cigarette advertising near schools? Prev Med. 2008;47: Chan WC, Leatherdale ST. Tobacco retailer density surrounding schools and youth smoking behaviour: a multi-level analysis. Tob Induc Dis. 2011;9: Leatherdale ST, Strath JM. Tobacco retailer density surrounding schools and cigarette access behaviors among underage smoking students. Ann Behav Med. 2007;33: Schneider JE, Reid RJ, Peterson NA, Lowe JB, Hughey J. Tobacco outlet density and demographics at the tract level of analysis in Iowa: implications for environmentally based prevention initiatives. Prev Sci. 2005;6: Siahpush M, Jones PR, Singh GK, Timsina LR, Martin J. Association of availability of tobacco products with socio-economic and racial/ethnic characteristics of neighbourhoods. Public Health. 2010;124: Wakefield M, Germain D, Durkin S, Henriksen L. An experimental study of effects on schoolchildren of exposure to point-of-sale cigarette advertising and pack displays. Health Educ Res. 2006;21: Wilkinson D, Abraham C. Constructing an integrated model of the antecedents of adolescent smoking. Br J Health Psychol. 2004;9: White V, Bayly M. Tobacco use by Victorian secondary school students in 2008 and trends over time. In: White V, Smith G, editors. Victorian Secondary School Students Use of Licit and Illicit Substances in 2008: Results from the 2008 Australian Secondary Students Alcohol and Drug Survey [Internet]. Melbourne (AUST): Victorian Department of Health; p [cited 2013 Aug 6]. Available from: Australian Bureau of Statistics Information Paper: An Introduction to Socio-Economic Indexes for Areas (SEIFA) Canberra (AUST): ABS; Hoving C, Reubsaet A, de Vries H. Predictors of smoking stage transitions for adolescent boys and girls. Prev Med. 2007;44: de Vries H, Candel M, Engels R, Mercken L. Challenges to the peer influence paradigm: results for year olds from six European countries from the European Smoking Prevention Framework Approach study. Tob Control. 2006;15: Dalglish E, McLaughlin D, Dobson A, Gartner C. Cigarette availability and price in low and high socioeconomic areas. Aust N Z J Public Health. 2013;37: Chapman S, Freeman B. Regulating the tobacco retail environment: beyond reducing sales to minors. Tob Control. 2009;18: Institute of Medicine. Ending the Tobacco Problem: A Blueprint for the Nation. Washington (DC): National Academies Press; AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 2013 vol. 37 no. 6