Smoking susceptibility among never-smokers: data from the 2006-07 national youth smoking survey.
Yang, Wiworn Sae ; Leatherdale, Scott T. ; Ahmed, Rashid 等
Starting smoking before 15 years of age is associated with
increased risk of lung cancer, (1) and increased risk of breast cancer
among females. (2) Despite the health risks of smoking, 14% of Canadian
youth aged 15 to 19 were still considered current smokers according to
the 2009 Canadian Tobacco Use and Monitoring Survey (CTUMS).3 Preventing
youth smoking onset must remain a public health priority in Canada.
One way to prevent youth smoking uptake is to target prevention
activities to the non-smoking youth who are at greatest risk for
becoming future smokers. (4) The concept of smoking susceptibility
(defined as a lack of a firm commitment not to smoke among
never-smokers) can be used to identify those at-risk youth. (5) Evidence
from various prospective studies suggested that non-smoking youth who
are susceptible to smoking are over twice as likely to experiment with
cigarettes compared to those who are nonsusceptible. (6,7) Considering
smoking susceptibility has been shown to be modifiable through
interventions, (8) identifying susceptible non-smoking youth and
understanding the factors associated with susceptibility is important
for improving future smoking prevention programs. (9)
According to the Theory of Triadic Influence,10 smoking
susceptibility can be influenced by factors from different levels of
context: individual characteristics (e.g., sex, other risk behaviours),
immediate social environment (e.g., friends and family) and broader
social environment (e.g., school). Research has previously identified
that characteristics from each of these levels of context are associated
with susceptibility. For example, non-smoking youth are more likely to
be susceptible if they are female, (11) or if they have friends or
family members who smoke. (12,13) Considering that research has
identified that youth smokers are more apt to use alcohol and/or
marijuana (14) and to be exposed to smoking in homes or in cars, (15) it
may also be informative to examine these characteristics in relation to
susceptibility. Similarly, considering that research has identified that
high-risk school environments (i.e., schools with a high prevalence of
smoking) are associated with smoking onset, (4,16,17) it may also be
worthwhile to examine this school characteristic in relation to
susceptibility. Thus, the purpose of this study is to examine the
association between school and individual characteristics and smoking
susceptibility of non-smoking youth.
METHODS
Design
This study used nationally representative data collected from
41,886 grade 9 to 12 students as part of the 2006-07 Canadian Youth
Smoking Survey (YSS). (18) In brief, the target population consisted of
all young Canadian residents in grades 9 to 12 attending public and
private secondary schools in 10 Canadian provinces. The YSS survey was
administered to students during class time and participants were not
provided compensation. Active information with passive consent was used
to reduce demands on schools and to increase student participation
rates. Detailed information on the 2006-07 YSS is available in print
(18) or online (www.yss.uwaterloo.ca).
Measures
Susceptibility among never-smokers was measured by asking students:
"Do you think in the future you might try smoking
cigarettes?", "If one of your best friends were to offer you a
cigarette, would you smoke it?", and "At any time during the
next year, do you think you will smoke a cigarette?" Consistent
with Pierce and colleagues, (5) students who answered 'definitely
not' to all three questions were considered non-susceptible; they
were considered susceptible to future smoking if they responded
positively to at least one item. Respondents were also asked to report
their sex, grade, if they have a parent or guardian who smokes, if they
have a sibling who smokes, how many of their five closest friends smoke,
the number of days they were exposed to smoking in a car during the
previous week, the number of people who smoke inside their home daily,
if they have ever tried marijuana, and if they have ever tried alcohol.
For each school, we also calculated the school-level prevalence of
students considered current smokers, where current smokers were defined
as those who have smoked at least 100 cigarettes in his/her lifetime and
have smoked in the 30 days preceding the survey.
Analyses
Descriptive analyses examining smoking susceptibility were
performed according to demographic characteristics, social influences,
substance use, and province. Since students (level 1) are nested within
schools (level 2), we then conducted three different multilevel logistic
regression models to examine school- and student-level factors
associated with the odds of a never-smoker being susceptible to smoking.
The first model was a null model to determine the school-level variance
([[sigma].sup.2.sub.[mu]0]) in the likelihood of being a susceptible
never-smoker. The second model examined if susceptibility was associated
with the school-level smoking rate and student-level social influences,
controlling for sex and grade. The third model examined if student-level
alcohol and marijuana use were associated with the odds of being
susceptible, controlling for sex, grade and the school-level smoking
rate. Survey weights were used in the descriptive analyses to adjust for
non-response between provinces and groups, thereby minimizing any bias
in the analyses caused by differential response rates across regions or
groups. The statistical package SAS 8.02 was used for all analyses. (19)
RESULTS
In 2006, 10.2% (167,921) of Canadian youth in grades 9 to 12 were
current smokers, 1.6% were former smokers, and 88.2% (1,452,041) were
never-smokers. The prevalence of never-smoking was higher for females
(89.9%) compared to males (86.7%) ([chi square] = 71.6, df=2,
p<0.001) and students in grades 11 and 12 were less likely to be
never-smokers than students in grades 9 and 10 ([chi square] = 497.1,
df=6, p<0.001). Among the never-smokers, 28.2% (255,840) were
considered to be susceptible to smoking and 71.8% (649,889) were
considered non-susceptible to smoking. The mean school smoking rate was
11.5% (range, 0% to 53.8%). Descriptive statistics by smoking
susceptibility among never-smokers are presented in Table 1.
Females were more likely than males to be susceptible to smoking
([chi square] = 17.2, df=1, p<0.001). Students in higher grades were
less likely to be susceptible to smoking than students in lower grades
([chi square] = 203.4, df=3, p<0.001). Youth with no close friends
were less likely to be susceptible than youth with smoking friends ([chi
square] = 275.4, df=5, p<0.001). Youth who have ever tried marijuana
([chi square] = 145.0, df=1, p<0.001) or alcohol ([chi square]=701.4,
df=1, p<0.001) were more likely to be susceptible to smoking than
youth who have not tried alcohol or marijuana. As illustrated in Figure
1, youth in Quebec and British Columbia were more apt to be susceptible
to smoking than youth in Atlantic Canada, Ontario or the Prairies ([chi
square] = 95.9, df=4, p<0.001).
Significant between-school random variation in smoking
susceptibility was identified [[[sigma].sup.2.sub.[mu]0]=0.18 (0.03),
p<0.001]; school-level differences accounted for 5.2% of the
variability in the odds of being considered susceptible to smoking. This
suggests that characteristics of the school that a non-smoking student
attends are associated with his/her risk of being susceptible to smoking
in the future.
The adjusted odds ratios examining if school- and student-level
social influences are associated with smoking susceptibility are
presented in Table 2. The odds of a never-smoker being susceptible to
smoking increased if the student had a sibling who smokes (OR 1.58, [95%
CI 1.35-1.85]), or if the student had 1 or 2 close friends (1.92,
[1.67-2.20]) or 3 or more close friends (1.75, [1.47-2.09]) who smoke.
Conversely, the odds of a never-smoker being susceptible to smoking
decreased if the student was male (0.83, [0.74-0.93]), or in grade 10
(0.79, [0.69-0.91]), grade 11 (0.64, [0.55-0.75]) or grade 12 (0.50,
[0.42-0.58]). Parental smoking, exposure to smoking in a car or in the
home, and the prevalence of smoking at the school were not significantly
associated with the odds of being susceptible to smoking.
As shown in Table 3, even when controlling for sex and grade, the
odds of a never-smoker being susceptible to smoking increased if the
student had ever used marijuana (1.43, [1.33-1.55]) or alcohol (2.61,
[2.41-2.82]) relative to students who have never used marijuana or
alcohol.
DISCUSSION
Smoking susceptibility is a strong predictor of smoking initiation
in youth as it is an early precursor in the sequence of cognitive
changes that ultimately leads to experimentation with smoking. (5-7) As
such, the measure of susceptibility can be used to identify high-risk
non-smoking youth before smoking onset occurs. One way to prevent
smoking onset among youth is to interfere with the development of
susceptibility to smoking in the non-smoking youth population. In this
study we identified that in 2006, more than 1 in 4 Canadian youth who
were never-smokers were considered susceptible to future smoking;
representing over 255,000 Canadian youth in grades 9 to 12. This
suggests that ongoing tobacco control primary prevention initiatives are
crucial for preventing smoking onset among non-smoking Canadian youth.
Consistent with previous research, (11) we identified that smoking
susceptibility among never-smokers varied significantly across schools.
However, unlike previous research that had identified an association
between the school-level prevalence of smokers at a school and the risk
of smoking onset and progression, (16,17) we did not identify an
association between the school-level prevalence of smoking at a school
and the likelihood of a never-smoker at that school being susceptible.
Although it cannot be determined with these cross-sectional data, it may
be due to the strong association between close-friend smoking and
smoking susceptibility, (7,11,20) or it may also be due to the impact
that school-level smoking social models may have on smoking onset
(4,16,17) (i.e., students who may have been susceptible had already
tried smoking due to their high-risk school environment). Regardless,
our finding that significant between-school variability exists suggests
that additional research is required to identify the school-level
characteristics that explain this variability. Such insight would be
valuable for developing school-based interventions to create school
environments that do not promote smoking onset.
As expected, non-smoking youth with smoking friends or smoking
siblings were more likely to be susceptible to future smoking than those
with no smoking friends. However, we did not find a significant
association between parental smoking and exposure to smoking in the home
and cars and susceptibility to smoking. As suggested by previous
research, youth appear more motivated to comply with the behavioural
norms of their peer group rather than of their parents. (21) For
instance, smoking friends can expose youth to the immediate positive
social outcomes associated with smoking, (12) making them more apt to
consider trying smoking. (22) Consistent with research on more advanced
smoking behaviour, (14,23) we also identified that smoking
susceptibility was associated with ever using alcohol and/or marijuana.
Considering these are also behaviours which often occur within peer
groups, it may be important future research to develop a better
understanding of the link between smoking onset and alcohol and
marijuana use, and to examine the benefits of using a more comprehensive
multi-substance approach to youth smoking prevention programming. (14)
Age and sex were also associated with smoking susceptibility.
Consistent with previous studies, (12,13) we found that females were
more likely to be susceptible to smoking than males. Although it cannot
be determined with these cross-sectional data, it is likely related to
the increased prevalence of young females starting to smoke compared to
males. (24) Likewise, students in older grades were less apt to be
susceptible to smoking, probably as a function of smoking onset
occurring as students move to higher grades. (24)
Limitations
Although the response rate was high and the data were weighted to
help account for non-response, the findings are nevertheless subject to
sample bias. In addition, the findings likely reflect some
under-reporting for tobacco use, as is common in survey research. (25)
It should also be noted that the cross-sectional nature of the design
does not allow for causal inferences. Data pertaining to school-level
tobacco control prevention programs and policies were not available,
limiting our ability to identify the characteristics of schools that are
associated with increased susceptibility.
CONCLUSION
Ongoing monitoring of smoking behaviours of students and
school-level characteristics will help researchers to identify at-risk
youth as well as schools that create high-risk environments for smoking
onset. Considering the large number of Canadian youth considered at high
risk for becoming future smokers, our results suggest that ongoing
tobacco control primary prevention initiatives are crucial for reducing
the prevalence of smoking susceptibility and preventing smoking onset
among non-smoking Canadian youth.
Acknowledgements: We thank the Propel Centre for Population Health
Impact for providing support for this project. The 2006-2007 Youth
Smoking Survey is a product of a pan-Canadian capacity-building project
that includes Canadian researchers from all provinces and provides
training opportunities for university students at all levels. Production
of this paper has been made possible through a financial contribution
from Health Canada. The views expressed herein do not necessarily
represent the views of Health Canada. Dr. Leatherdale is a Cancer Care
Ontario Research Chair in Population Studies funded by the Ontario
Ministry of Health and Long-Term Care.
Conflict of Interest: None to declare.
Received: December 10, 2010
Accepted: March 3, 2011
REFERENCES
(1.) Peto R, Darby S, Deo H, Silcocks P, Whitley E, Doll R.
Smoking, smoking cessation and lung cancer in the UK since 1950:
Combination of nutritional statistics with two case-control studies.
BMJ2000;321:323-29.
(2.) Band PR, Le ND, Fang R, Deschamps M. Carcinogenic and
endocrine disrupting effects of cigarette smoke and risk of breast
cancer. Lancet 2002;360:1044-49.
(3.) Health Canada. Canadian Tobacco Use Monitoring Survey, Wave 1
Results 2009. Ottawa, ON: Health Canada, 2010. Available at:
http://www.hcsc.gc.ca/hc-ps/tobac-tabac/research-recherche/stat/
_ctums-esutc_2009/w-p1_sum-som-eng.php (Accessed Sept 22, 2010).
(4.) Cameron R, Brown KS, Best JA, Pelkman CL, Madill CL, Manske
SR, et al. Effectiveness of a social influences smoking prevention
program as a function of provider type, training method, and smoking
risk. Am J Public Health 1999;89(12):1827-31.
(5.) Pierce JP, Choi WS, Gilpin EA, Farkas AJ. Validation of
susceptibility as a predictor of which adolescents take up smoking in
the United States. Health Psychol 1996;15(5):355-61.
(6.) Jackson C. Cognitive susceptibility to smoking and initiation
of smoking during childhood: A longitudinal study. Prev Med
1998;27(1):129-34.
(7.) Unger JB, Johnson CA, Stoddard JL, Nezami E, Chou CP.
Identification of adolescents at risk for smoking initiation: Validation
of a measure of susceptibility. Addict Behav 1997;22(1):81-91.
(8.) Meschack AF, Hu S, Pallone UE, McAlister AL, Gottlieb N, Huang
P. Texas tobacco prevention pilot initiative: Process and effects.
Health Educ Res 2004;19(6):657-68.
(9.) Wilkinson AV, Waters AJ, Vasudevan V, Bondy ML, Prokhorov AV,
Spitz MR. Correlates of susceptibility to smoking among Mexican origin
youth residing in Houston, Texas: A cross-sectional analysis. BMC Public
Health 2008;8:337.
(10.) Flay BR, Petraitis J, Hu FB. Psychosocial risk and protective
factors for adolescent tobacco use. Nicotine Tob Res 1999;1:S59-S65.
(11.) Leatherdale ST, Brown SK, Cameron R, McDonald PW. Social
modeling in the school environment, student characteristics, and smoking
susceptibility: A multi-level analysis. J Adolesc Health
2005;37(4):330-36.
(12.) Feighery E, Borzekowski DL, Schooler C, Flora J. Seeing,
wanting, owning: The relationship between receptivity to tobacco
marketing and smoking susceptibility in young people. Tob Control
1998;7(2):123-28.
(13.) Kaufman NJ, Castrucci BC, Mowery PD, Gerlach KK, Emont S,
Orleans T. Predictors of change on the smoking uptake continuum among
adolescents. Arch Pediatr Adolesc Med 2003;156:581-87.
(14.) Leatherdale ST, Ahmed R. Alcohol, marijuana, and tobacco use
among Canadian youth: Do we need more multi-substance prevention
programming? J Prim Prev 2010;31(3):99-108.
(15.) Leatherdale ST, Ahmed R. Second-hand smoke exposure in homes
and in cars among Canadian youth: Current prevalence, beliefs about
exposure, and changes between 2004 and 2006. Cancer Causes Control
2009;20(6):855-65.
(16.) Leatherdale ST, Manske SR. The relationship between student
smoking in the school environment and smoking onset in elementary school
students. Cancer Epidemiol Biomarkers Prev 2005;14(7):1762-65.
(17.) Leatherdale ST, Cameron R, Brown SK, McDonald PW. Senior
student smoking at school, student characteristics, and smoking onset
among junior students: A multilevel analysis. Prev Med
2005;40(6):853-59.
(18.) Health Canada. 2006-07 Youth Smoking Survey, Ottawa: Minister
of Supply and Services Canada, 2008. Available at:
http://www.hc-sc.gc.ca/hl-vs/tobactabac/research-recherche/stat/survey-
sondage_term-eng.php (Accessed October 30, 2008).
(19.) SAS Institute Inc. The SAS system for windows. Cary, NC: SAS
Institute Inc., 2001.
(20.) Aloise-Young PA, Graham JW, Hansen WB. Peer influence on
smoking initiation during early adolescence: A comparison of group
members and group outsiders. J Appl Psychol 1994;79(2):281-87.
(21.) Harris JR. Where is the child's environment? A group
socialization theory of development. Psychol Rev 1995;102(3):458-89.
(22.) Bandura A. Social Foundations of Thought and Action: A Social
Cognitive Theory. Englewood Cliffs, NJ: Prentice Hall, 2001.
(23.) Bobo JK, Husten C. Sociocultural influences on smoking and
drinking. Alcohol Res Health 2000;24(4):225-32.
(24.) Health Canada. Canadian Tobacco Use Monitoring Survey, Wave 1
Results 2003. Ottawa: Health Canada, 2004.
(25.) Bovet P, Viswanathan B, Faeh D, Warren W. Comparison of
smoking, drinking, and marijuana use between students present or absent
on the day of a school-based survey. J Sch Health 2006;76(4):133-37.
Correspondence: Scott T. Leatherdale, PhD, Department of Prevention
and Cancer Control, Cancer Care Ontario, 620 University Avenue, Toronto,
ON M5G 2L7, Tel: 416971-9800, ext. 3237, Fax: 416-971-7554, E-mail:
scott.leatherdale@cancercare.on.ca
Wiworn Sae Yang, BSc, [1] Scott T. Leatherdale, PhD, [2,3] Rashid
Ahmed, PhD [3]
Author Affiliations
[1.] Department of Health Studies & Gerontology, University of
Waterloo, Waterloo, ON
[2.] Department of Prevention and Cancer Control, Cancer Care
Ontario, Toronto, ON
[3.] Propel Centre for Population Health Impact, Canadian Cancer
Society and the University of Waterloo, Waterloo, ON
Table 1. Descriptive Statistics for the Sample of Never-
smoking Youth in Grades 9 to 12 by Smoking
Susceptibility Status, Canada, 2006
Demographics Susceptible
Never-smokers
(n=255,840) *
%
Sex Male 48.4
Female 51.6
Grade 9 36.5
10 28.3
11 21.3
12 14.0
Social Influences
Parent/guardian smokes Yes 36.0
No 64.0
Sibling smokes Yes 15.8
No 84.2
Number of close friends None 62.0
who smoke 1 14.8
2 9.9
3 3.9
4 2.6
[greater than 6.8
or equal to] 5
Number of days exposure 0 days 78.8
to smoking in a car during 1-6 days 19.6
the previous week All 7 days 1.6
Number of people smoking None 83.5
inside the home daily 1 person 10.1
[greater than 6.4
or equal to]
2 people
Substance Use
Ever tried marijuana Yes 19.4
No 80.6
Ever tried alcohol Yes 79.3
No 20.7
Demographics Non-susceptible
Never-smokers
(n=649,889) *
Sex Male 51.4
Female 48.6
Grade 9 28.5
10 26.4
11 23.5
12 21.6
Social Influences
Parent/guardian smokes Yes 32.9
No 67.1
Sibling smokes Yes 10.9
No 89.2
Number of close friends None 74.2
who smoke 1 11.3
2 6.3
3 3.1
4 1.1
[greater than 4.0
or equal to] 5
Number of days exposure 0 days 81.3
to smoking in a car during 1-6 days 16.9
the previous week All 7 days 1.8
Number of people smoking None 85.3
inside the home daily 1 person 9.0
[greater than 5.7
or equal to]
2 people
Substance Use
Ever tried marijuana Yes 12.7
No 87.3
Ever tried alcohol Yes 61.7
No 38.3
* Population estimate.
Table 2. Logistic Regression Analyses Examining Social
Influences Associated With Smoking Susceptibility
Among Never-smokers in Grades 9 to 12, Canada,
2006
Adjusted Odds Ratio ([dagger])
(95% CI)
Model 1
Susceptible vs.
Non-susceptible
Student-level
Characteristics
Sex Female 1.00
Male 0.82 (0.74-0.92) *
Grade 9 1.00
10 0.79 (0.69-0.91) *
11 0.64 (0.55-0.74) *
12 0.49 (0.42-0.58) *
Parent/guardian smokes No 1.00
Yes 0.99 (0.86-1.13)
Sibling smokes No 1.00
Yes 1.58 (1.35-1.85) *
Number of close None 1.00
friends who smoke 1 or 2 1.91 (1.67-2.19) *
[greater than 1.76 (1.47-2.10) *
or equal to] 3
Number of days All 7 days 1.00
exposure to smoking
in a car during the 1-6 days 1.42 (0.88-2.28)
previous week 0 days 1.28 (0.80-2.06)
Number of people smoking None 1.00
inside the home daily 1 person 1.23 (1.00-1.50)
[greater 0.89 (0.69-1.15)
than or people
equal to] 2
School-level Characteristic
School smoking rate Each 1% 0.22 (0.04-1.37)
increase
([dagger]) Odds ratios adjusted for all other variables in the table.
* p<0.001.
Table 3. Logistic Regression Analyses Examining Substance
Use Associated With Smoking Susceptibility Among
Never-smokers in Grades 9 to 12, Canada, 2006
Adjusted Odds Ratio ([dagger])
(95% CI)
Model 2
Susceptible vs.
Non-susceptible
Student-level Characteristics
Ever use marijuana No 1.00
Yes 1.44 (1.32-1.57) *
Ever use alcohol No 1.00
Yes 2.61 (2.42-2.81) *
([dagger]) Odds ratios adjusted for all other variables in the table
and controlling for sex, grade and the school smoking rate.
* p<0.001.
Figure 1. Smoking susceptibility among never-smokers in
grades 9-12 by region of Canada, 2006-2007
Susceptible Non-Susceptible
Atlantic ([dagger]) 24.4 75.6
Quebec 32.8 67.2
Ontario 27.8 72.2
Prairies ([double dagger]) 26.5 73.5
British Columbia 29.4 70.6
Note: Table made from bar graph.
