JOURNAL OF ADOLESCENT HEALTH 2000;27:281–288
INTERNATIONAL ARTICLE
Cigarette Smoking in Multicultural, Urban High School Students PHILIPPA HOLOWATY, M.Sc., Ph.D., LINDA FELDMAN, R.N., M.Sc.N., BART HARVEY, M.D., Ph.D., AND LINDA SHORTT, R.N., M.Ed.
Purpose: To profile patterns of cigarette use among a multiethnic population of high school students, and identify important factors associated with cigarette use by ethnicity, in order to plan effective health promotion strategies. Methods: This cross-sectional study involved the completion of a lifestyle questionnaire by 1236 Grade 9 –13 students (86% response rate) from 62 randomly selected classrooms in three urban high schools in Toronto. Chisquare analysis of the association between tobacco use and other variables took account of the clustered sample using CSAMPLE in Epi Info. Results: The students self-identified their ethnicity as follows: 388 Canadian, 269 European, 171 East Indian, 137 Asian, 76 West Indian, and 194 “other.” Students who identified themselves as Canadians were significantly more likely to be current smokers (29%) than students reporting other ethnicities (13%). There was no apparent increase in smoking rates for immigrants after 2 or more years in Canada. Current drinking, sexual activity, and especially friends smoking was most strongly associated with current smoking for most ethnic groups, although the relative importance of these variables was not identical for all groups. Conclusions: Prevention programs may benefit from a focus on the influence of peer smoking and on the grouping together of lifestyle factors associated with smoking for students in all ethnic groups in this multicultural city. © Society for Adolescent Medicine, 2000
From Toronto Public Health, Toronto, Ontario, Canada Address correspondence to: Linda Feldman, R.N., M.Sc.N., Toronto Public Health, 590 Jarvis Street, 3rd Floor, Toronto, Ontario M4Y 2J4, Canada. Manuscript accepted October 15, 1999.
KEY WORDS: Adolescence Acculturation Smoking Peer group Ethnicity Risk behaviors Canada
Smoking is the leading preventable cause of death. It kills an estimated 13,000 Ontario residents each year through its contribution to heart disease, lung and other cancers, chronic lung disease, and stroke (1). Most smokers in North America begin their addiction to smoking in their teens (2). There is concern that cigarette smoking is not declining in teenagers despite smoking prevention programming (2). Most studies of smoking behaviors in youth have ignored cultural differences, and there is a lack of research into smoking prevalence and the influence of ethnicity and immigrant status in Canada (3,4). The prevalence and patterns of cigarette use among adolescents may vary among different ethnic groups and with acculturation. Several Canadian studies have found that smoking rates are lower in foreignborn residents, that immigrants increase cigarette consumption after arriving in North America, and that the proportion of immigrants who are smokers rises with increased time spent in the new country (4 –7). Previously identified predictors and correlates of teenage smoking may not be relevant or may differ in the degree of importance across ethnic groups for adolescents (4). The present study took place in the Borough of East York, one of the six municipalities of Metropol-
© Society for Adolescent Medicine, 2000 Published by Elsevier Science Inc., 655 Avenue of the Americas, New York, NY 10010
1054-139X/00/$–see front matter PII S1054-139X(99)00122-6
282
HOLOWATY ET AL.
itan Toronto, Ontario. It is an urban area with a multicultural population of just more than 100,000, 20% of whom are under age 20 years (8). A 1991 survey of Toronto high school students reported that 43% of students were born outside of Canada and almost half were native speakers of a language other than English or French (9). The ethnic mix in Toronto high schools varies from one area to another, and Health Unit staff sensed that patterns of health behaviors also varied. In December 1994, a study was undertaken to identify specific behaviors among local high school students (10); to determine whether relationships exist between lifestyle beliefs and behaviors [tobacco use, safety, alcohol use (11), physical activity (12), sexual activity (13), and family functioning] and various sociodemographic factors (gender, grade, parent education, birthplace, and religious attendance); and to assist in the development and implementation of health promotion initiatives. This report focuses on the tobacco use section of the questionnaire, which included questions on smoking behavior, and smoking cessation.
Methods This was a cross-sectional survey using a self-administered questionnaire completed by students in homeroom classes. The sample size was calculated using the STATCAL program, which is part of Epi Info software (14). A random cluster sample was chosen by grade and homeroom from class lists submitted by the principals at the three selected high schools. Sixty-two of a possible 183 homerooms were selected for this study. Ethical approval was received from both the East York Board of Health and the East York Board of Education. Four weeks before survey administration, a notice was given to the selected homeroom teachers and students explaining the survey and stating that participation was voluntary. One week later, a similar letter was sent to parents of the students. Parents who had questions, or did not wish their children to take part, were asked to call the Director of Adolescent Health at the Health Unit. On the day of survey administration, trained surveyors provided an explanation of the study to the students, emphasizing that their participation was voluntary and that the survey was completely confidential. The 125-item questionnaire was constructed from questions drawn from 17 other surveys (list available upon request) and from questions arising from practice. Each section of the questionnaire was assessed
JOURNAL OF ADOLESCENT HEALTH Vol. 27, No. 4
by experts for face validity. A pilot test of the methods and questionnaire was carried out with 40 students enrolled in a local high school summer program. Minor changes were made to clarify the meaning of certain questions for this multiethnic population. The questions in the tobacco use section were developed from questions adapted from the Tobacco Free High School Survey, 1990 (15), Smoke Free Nova Scotia, 1994 (16), Health of Canada’s Youth, 1992 (17), the University of Waterloo Lifestyle questionnaire, 1987 (18), the Ontario Health Survey, 1990 (19), and the Thunder Bay Health Unit Survey, 1993 (20). Psychometrics for three of these surveys (17–19) have been described but are not available for the three other surveys (15,16,20). The aim was to make the questionnaire relevant to future program planning and as comparable as possible to other wellestablished surveys. The smoking profile of students was obtained by combining information from tobacco use Questions 6, 9, and 10. Four main categories were used to classify students’ smoking status: never smoked (has not had even one puff), not a current smoker (includes those who have tried only one cigarette, and quitters), occasional smoker (currently smokes but not on a daily basis), and daily smoker (has at least one cigarette each day). However, for testing associations (relative risks [RRs]), smoking was treated as a bivariate outcome with occasional and daily smokers classified together as current smokers versus other students. Students were asked separately about their father’s/male guardian’s and mother’s/female guardian’s smoking habits. The variable Parental Smoking reflects the greatest level of smoking in either parent, or the level of smoking in the one parent when information was available only for one parent. The family functioning scale was as used in the Ontario Health Survey (1992) (19). The scale was originally taken from the general functioning subscale (GL) of the McMaster Family Assessment Device (FAD) that was used in the Ontario Child Health Study (21). The GL is a 12-question shorter version and correlates highly with the longer FAD. The cutoff point marking healthy from dysfunctional families has been identified as that best distinguishing families seeking clinical help from those in the general population (21). Ethnicity was derived from the answers to the following question: “Please write down the term that best describes the ethnic character of your everyday home environment (for example: Canadian, West Indian, East Indian, Chinese, Greek, etc.).” Ethnici-
October 2000
MULTICULTURAL ADOLESCENT SMOKING
283
Table 1. Description of Sample by Stated Ethnicity Canadian
Grade 9 10 11 12 13 Gender Male Female Years since immigration Not stated ⬍2 y 2-4 y ⱖ5 y Born in Canada Parent education Not stated Attended university Graduated HS No HS graduate Total %
European
East Indian
Asian
West Indian
Other
Total
Col %
CI
Col %
CI
Col %
CI
Col %
CI
Col %
CI
Col %
CI
19 25 21 13 22
(11–27) (14 –36) (13–29) (8 –18) (12–33)
12 26 27 18 17
(7–17) (13– 40) (15–39) (11–25) (9 –24)
13 25 29 10 22
(5–20) (9 – 42) (13– 46) (4 –16) (9 –36)
17 27 26 11 18
(8 –27) (13– 41) (14 –37) (5–18) (4 –32)
22 17 28 17 16
(12–33) (6 –28) (13– 42) (8 –25) (1–33)
23 28 25 10 15
(13–33) (16 –39) (16 –34) (6 –15) (7–23)
213 314 310 162 237
17 25 25 13 19
52 48
(45–56) (41–56)
51 49
(42– 60) (40 –59)
46 54
(32–59) (41– 68)
49 51
(40 –58) (42– 60)
46 54
(35– 60) (41– 68)
57 43
(49 – 65) (35–51)
628 608
51 49
2 2 2 3 92
(0 –3) (0 –3) (1–5) (90 –95)
0 5 5 6 84
(2– 8) (2– 8) (4 –9) (77– 89)
2 12 12 32 42
(6 –19) (7–16) (25– 40) (15–31)
4 12 18 29 37
(2–21) (10 –26) (22–36) (28 – 46)
4 13 10 18 55
(5–21) (3–18) (5–27) (42– 67)
11 23 19 20 28
(15–31) (12–26) (12–27) (20 –36)
42 112 108 175 799
3 9 9 14 65
(31– 48) (26 – 40) (1–7)
173 479 450 134
14 39 36 11 100
7 55 32 7 31
(49 – 61) (26 –37) (3–11)
13 23 37 28 22
(16 –29) (31– 44) (21–34)
12 36 45 7 14
(29 – 43) (37–53) (3–11)
ties were then grouped into six categories arising from the data: Canadian, European (both Eastern and Western Europe), Asian (Chinese, Japanese, Vietnamese, Philippine, Korean, and Sri Lankan), East Indian, West Indian, and other countries. The single ethnic identity students chose was a measure of acculturation to the Canadian lifestyle. Birthplace was categorized as “Canadian born” or “not Canadian born.” Students who were not born in Canada were asked in which year they arrived. Analysis Each survey represented one student, but surveys were given a weight of ⬍ 1 ⬍ if students from the particular school and grade had been under- or overrepresented by the sampling method. The data were analyzed by simple frequencies, cross-tabulations, stratified analysis, and multivariate logistic regression using SPSS for Windows (22). As the students had been sampled by class rather than individually, allowance had to be made for this in calculating the precision of estimated percentages and RRs. This was done using the CSAMPLE program within Epi Info software (14). It was not possible to account for cluster sampling in logistic regression using these software packages. Ignoring cluster sampling may lead to associations appearing
21 37 33 8 11
(26 – 49) (26 – 41) (3–14)
20 23 53 4 6
(13–33) (41– 66) (0 – 8)
24 39 33 4 16
Count Col %
significant when they are not. We therefore used a conservative cutoff p value ⱕ .01 in the logistic regression.
Results A total of 1434 students from 62 classes in three high schools were invited to participate. Fourteen parents called with questions about the survey and seven asked that their children be excluded. On the day of survey administration, 1245 (87%) were present. Nine inconsistent or incomplete questionnaires were excluded, leaving a total of 1236 for analysis. The 1236 students surveyed were 28% of the total 4361 students registered in the three high schools. The distribution of the weighted sample by ethnicity and sociodemographic characteristics is given in Table 1. Students who identified their ethnicity as Canadian were mostly born in Canada (94%), as were students whose stated ethnicity was European (84%). The students with “other” ethnicities were from many different countries and 23% had arrived within the past 2 years. The overall prevalence of smoking (occasional and daily smokers) for the high school students was 18% [95% confidence interval (CI), 15–19%]. Canadian and European students tended to smoke daily more than the other groups (Table 2). Sixty percent or
284
HOLOWATY ET AL.
JOURNAL OF ADOLESCENT HEALTH Vol. 27, No. 4
Table 2. Description of Sample by Smoking Status Never Smoked
Born in Canada? Not stated No Yes Years since immigration Not stated ⬍2 y 2– 4 y ⬎5 y Born in Canada Stated ethnicity Canadian European East Indian Asian West Indian Other Total
Not Current Smoker
Occasional Smoker
Row %
(CI)
Row %
(CI)
Row %
(CI)
Row %
(CI)
Count
Col %
78 59 40
(54 – 65) (34 – 46)
9 31 38
(25–36) (33– 43)
13 5 7
(3– 8) (4 –9)
0 5 15
(2–7) (12–18)
11 425 799
1 34 65
65 71 56 53 40
(60 – 81) (47– 65) (46 – 60) (34 – 46)
19 22 31 37 38
(12–32) (22– 41) (30 – 45) (33– 43)
9 3 6 6 7
(0 –7) (2–10) (2–10) (4 –9)
7 4 6 3 15
(0 –7) (2–11) (1– 6) (13–18)
42 112 108 175 799
3 9 9 14 65
32 40 66 63 40 63 47
(25–39) (33– 48) (57–75) (54 –72) (26 –53) (56 –70) (43–52)
40 39 28 27 51 26 35
(33– 47) (32– 46) (20 –36) (19 –35) (37– 66) (20 –33) (31– 40)
7 9 3 8 6 4 6
(4 –9) (5–13) (0 –7) (3–13) (0 –12) (1–7) (5– 8)
22 12 3 2 3 6 11
(17–27) (7–16) (1–5) (0 –5) (0 – 6) (3–10) (9 –14)
388 269 171 137 78 194 1236
31 22 14 11 6 16 100
more of Canadian, European, and West Indian students had smoked, as opposed to 37% of Asian students, 34% of East Indian students, and 37% of “other” students. Although West Indian students experimented with cigarettes at the same rate as Canadian and European students, very few were daily smokers compared with these other two groups. In fact, very few students who identified themselves as being in the East Indian, West Indian, or Asian groups smoked daily. Birthplace in Canada was associated with higher percentages of daily smokers and decreased percentage of students who had never tried cigarettes (Table 2). There was no apparent increase in smoking rates for immigrants after 2 or more years in Canada. In logistic regression, Canadian ethnicity was a better predictor of current smoking than birthplace in Canada. Table 3 summarizes the univariate associations between current smoking and bivariate demographic and lifestyle variables. Students reporting their ethnicity as Canadian were significantly more likely to be current smokers than students reporting other ethnicities. Using multivariate logistic regression, the variables which remained significantly associated (p ⬎ .01) with current smoking status (and with daily smoking status) were: half or more of friends smoke, sexually active, and current drinker of alcohol (Table 4). Ethnicity (Canadian vs. “other”) did not remain a significant predictor of smoking status in the multivariate model. A higher percentage of Asian (55%), East Indian (64%), and “other” (46%) students had
Daily Smoker
Total
none of the previously stated behaviors (current smoking, sexual activity, alcohol use, and friends who smoke) compared with Canadian (22%), European (23%), and West Indian (24%) students. Table 5 shows the RR of being a current smoker associated with each variable of interest, separately for each ethnic group. When 50% or more of their friends smoked, students themselves were more likely to be current smokers for all ethnic groups. Students who reported themselves to be sexually active and/or current drinkers were also more likely to be current smokers, but the association was not statistically significant in all ethnic groups. Physical activity, family functioning, gender, having at least one parent who ever smoked, and contentment with life were not significantly associated with current smoking within ethnic groups. Table 4 shows the multivariate model for the three main factors associated with current smoking, by ethnicity. In the multivariate models there was a tendency for friends’ smoking to increase in importance and sexual activity and alcohol use to become less important than in univariate analysis. However, for Asians, sexual activity became the most important variable for explaining current cigarette use. The association between current smoking and current drinking was most important for Canadians and Europeans.
Discussion Our results reflect the somewhat lower adolescent smoking rates reported for Toronto than other cities.
October 2000
MULTICULTURAL ADOLESCENT SMOKING
285
Table 3. Current Smoking Status Associated With Psychosocial and Demographic Variables Variable Stated ethnicity Grade Gender Birthplace Happiness Family functioning Physical activity Sexual activity Alcohol Drunk driving (drivers only) Parents’ smoking Parents’ education Friends’ smoking
Not Canadian Canadian Grade 9 Grade 10 –13 Female Male Not Canada Canada Happy Not happy Healthy Dysfunctional Physically active Not physically active Not sexually active Sexually active Not current drinker Current drinker Never driven drunk Ever driven drunk Neither parent ever Either ever smoked High school grad Half or less smoke More than half smoke
No. of Students
% Current Smokers
RR*
848 388 213 1024 608 628 394 799 962 186 688 449 867 340 904 332 560 615 420 66 442 752 929 134 795 422
13 29 8 20 18 18 9 22 18 23 16 23 18 19 10 39 4 30 18 39 10 23 18 28 6 39
1.0 2.3 1.0 2.4 1.0 1.0 1.0 2.3 1.0 1.3 1.0 1.5 1.0 1.1 1.0 4.1 1.0 7.1 1.0 2.2 1.0 2.3 1.0 1.6 1.0 6.3
(95% CI) (1.7–3.2)* (1.6 –3.7)* (0.7–1.4) (1.6 –3.3)* (1.0 –1.9) (1.1–1.9)* (0.8 –1.5) (3.0 –5.6)* (4.5–11.1)* (1.5–3.2)* (1.6 –3.3)* (1.1–2.2)* (4.0 –10.0)*
Chi-square analysis accounts for cluster sampling. *Relative risks (RRs) are significantly different from 1.0.
Eighteen percent of students in our sample were current smokers (95% CI, 14 –21%). A smoking prevalence of of 24% was reported in the 1994 National Youth Smoking Survey (23). The Ontario Student Drug Use Survey organized by the Addiction Research Foundation (Metro Toronto Research Group on Drug Use, 1996) (24) reported that the proportion of students ages 12–19 years reporting tobacco use in 1995 in Metropolitan Toronto was 20% compared with 28% for the Province of Ontario. The lower smoking rate in our study may reflect the city’s high proportion of students from various ethnic groups who smoke very little. It is likely that there may have been a bias toward underreporting of smoking status (25), although reporting of current smoking status (smoker vs. nonsmoker) is fairly reliable (26). No attempt was made to verify smoking status as reported by the students. A recent similar survey of Toronto high school students reported that ethnicity remained a significant variable in a multivariate model of smoking status (4). However, their study did not account for sexual activity or alcohol use. In our study, the
addition of these two variables to the model resulted in all other variables becoming insignificant, except friends’ smoking. The difference in prevalence of smoking by ethnicity may be explained by the difference in prevalence of factors associated with smoking. Approximately half of students who identified themselves as East Indian, Asian, or “other” were not sexually active, did not smoke or drink alcohol, and had few friends who smoked. This survey did not collect information on academic grades, which have also been found to be strongly associated with smoking status (27), and which might also have been important in a multivariate model. Several studies have found higher prevalence of smoking amongst nonethnic North Americans, but this is not universally true (3). The degree of acculturation of the ethnic student also affected smoking behaviors (28,29). In our study the degree of acculturation, as measured by birthplace in Canada or, more important, as self-reported Canadian ethnicity, was related to smoking status in univariate analysis. However, uptake of smoking with years since immi-
286
HOLOWATY ET AL.
JOURNAL OF ADOLESCENT HEALTH Vol. 27, No. 4
Table 4. Multivariate Models of Association Between Three Main Risk Factors and Current Smoking for Each Ethnic Group Regression Model
N
Canadian* ⬎50% friends smoke Current drinker of alcohol Sexually active 2 ⫽ 173.808; 3 df; p ⬍ .00001* European current drinker* ⬎50% friends smoke Current drinker of alcohol Sexually active 2 ⫽ 37.781; 3 df; p ⬍ .00001* East Indian* ⬎50% friends smoke Current drinker of alcohol Sexually active 2 ⫽ 23.388; 3 df; p ⬍ .00001* Asian* ⬎50% friends smoke Current drinker of alcohol Sexually active 2 ⫽ 13.875; 1 df; p ⫽ .0031* West Indian ⬎50% friends smoke Current drinker of alcohol Sexually active 2 ⫽ 6.354; 3 df; p ⫽ .0956* Other* ⬎50% friends smoke Current drinker of alcohol Sexually active 2 ⫽ 25.185; 3 df; p ⬍ .00001* For total sample* ⬎50% friends smoke Current drinker of alcohol Sexually active 2 ⫽ 301.469; 3 df; p ⬍ .00001*
378
RR
p
8.1* 19.0* 4.7*
⬍.00001 ⬍.00001 ⬍.00001
3.1* 6.2* 1.7
⬍.0009 ⬍.0044 ⬍.1264
25.4* 4.5 1.4
⬍.0014 ⬍.1916 ⬍.7790
2.2 2.4 6.3*
⬍.2609 ⬍.2038 ⬍.0089
7.5 0.7 1.8
⬍.0273 ⬍.7059 ⬍.5063
8.0* 2.3 1.5
⬍.0003 ⬍.1560 ⬍.4892
5.7* 5.0* 2.7*
⬍.00001 ⬍.00001 ⬍.00001
259
159
128
77
166
1166
Logistic regression in SPSS, cluster sampling ignored. *Relative risk (RR) is significantly different from 1.0 (p value ⬍.01).
gration did not seem important, particularly after the first 2 years. This is contrary to several other reports on smoking uptake of immigrants to Canada; the reason for this difference is not apparent. Smoking behavior for recent immigrants will also be influenced by smoking practices in their native country. The prevalence of smoking in high school students in Korea in 1989 was found to be 8% (15% for boys and 1% for girls) (30), whereas smoking prevalence for high school students in Beijing, China, was reported to be 20% (29% for males and 11% for females) (31). A difference in smoking prevalence by gender has been reported for Asian Americans (32) and for Hispanic Americans (33). A study of smoking in U.S.-born black adults in contrast to West Indian– born Blacks in New York City found smoking rates
to be much lower and gender differences to be much greater in the foreign-born adults (34). In our study, there was no overall difference between male and female smoking rates, and differences in these rates within ethnic groups were not statistically significant. Expected gender differences within ethnic groups have not always been found (28), but this may be related to the level of acculturation of the groups. Smoking by friends was associated with current smoking in all ethnic groups in univariate analysis (Table 4). This was as expected from the literature, although the influence of peers has been reported to be less important for Blacks (28). In our study, the influence of peer smoking seemed less important for Asians. Smoking was also associated with alcohol
October 2000
MULTICULTURAL ADOLESCENT SMOKING
287
Table 5. Percentage of Current Smokers* Associated With Different Variables, by Ethnicity Variables (Low-Risk Group/ High-Risk Group)
Canadian RR†
European RR†
East Indian RR†
Asian RR†
West Indian RR†
Other RR†
⬍50% friends smoke ⬎50% friends smoke
7.6‡
(3.4 –16.7)
3.4‡
(1.8 – 6.3)
20.0‡
(2.3–100)
3.2‡
(1.0 –10.0)
5.9‡
(1.3–25.0)
7.1‡
(2.7–20.0)
Current nondrinker Current drinker
20.8‡
(5.6 –100)
6.9‡
(2.7–16.7)
4.1
(0.9 –20.0)
3.7‡
(1.3–11.1)
1.3
(0.3–5.0)
3.6‡
(1.6 – 8.3)
Not sexually active Sexually active
4.3‡
(2.9 – 6.7)
2.2‡
(1.3–36)
6.8‡
(1.8 –25.0)
6.6‡
(2.3–20.0)
2.1
(0.5–9.1)
2.6‡
(1.1–5.9)
Parents nonsmokers Parents ever smoke
1.8‡
(1.1–2.7)
2.0
(1.0 – 4.2)
1.7
(0.5–5.6)
2.5
(0.8 –7.7)
1.9
(0.5–7.1)
2.8
(0.9 – 8.3)
Functional family Dysfunctional family
1.4
(1.0 –2.0)
1.2
(0.7–2.0)
2.7
(0.6 –12.5)
1.9
(0.6 – 6.3)
4.2
(0.8 –20.0)
3.4‡
(1.1–10.0)
Female Male
0.8
(0.5–1.2)
1.0
(0.6 –1.5)
1.9
(0.4 –9.1)
1.5
(0.4 –9.1)
1.9
(0.5–7.7)
1.8
(0.7– 4.5)
Unhappy Happy
1.3
(0.9 –2.1)
1.0
(0.5–2.4)
1.2
(0.2–5.9)
2.3
(0.6 – 8.7)
1.2
(0.2–9.0)
2.9‡
(1.2–7.1)
Not physically active Physically active
1.4
(1.0 –2.2)
1.3
(0.7–2.2)
1.2
(0.3– 4.6)
1.1
(0.4 –3.3)
1.2
(0.4 –3.5)
1.1
(0.5–2.6)
Chi-square analysis accounting for clustered sampling Relative risk (RR) with 95% confidence interval in parentheses. ‡ Significantly different from 1.0. †
use and sexual activity within ethnic groups except West Indians. The association between smoking and use of other drugs including alcohol has been reported within ethnic groups by others (32,35). Our West Indian students were also anomalous because many of the students had tried cigarettes but were not current smokers. However, the numbers were small. We were unable to explore reasons for this anomaly in the cross-sectional study. It is recognized that ethnicity in this study may not accurately reflect the cultural diversity within the broader ethnic categories chosen. We also did not measure strength of ethnic identity. The study is an exploratory study with multiple tests of significance for the various associations. Further investigation is needed into the patterns of lifestyle behaviors associated with cigarette use by ethnicity and the implication of these patterns for health promotion. In summary, smoking was associated with having friends who smoked, and with alcohol use and sexual activity for the total sample and within most of the ethnic groups. Students who identified themselves as belonging to an ethnic group other than Canadian and European tended to be more recent immigrants and to have a lower prevalence of smoking and factors associated with smoking.
References 1. Schabas R. Opportunities for Health. Report of the Chief Medical Officer of Health of Ontario. Toronto: Queen’s Printer for Ontario, 1992:4 – 6. 2. Schabas R. Tobacco: Sounding the Alarm. Report of the Chief Medical Officer of Health of Ontario. Toronto: Queen’s Printer for Ontario, 1998. 3. Yang M. Ethnicity and Adolescent Smoking. The Ontario Tobacco Research Unit: Literature Reviews Series No. 9, Toronto, Ontario, Canada 1997. 4. Yang M, Skinner H. Dimensions of Ethnicity as Predictors of Adolescent Cigarette Smoking. The Ontario Tobacco Research Unit: Working Paper Series No. 41, Toronto, Ontario, Canada, 1998. 5. Millar WJ. Place of birth and ethnic status: Factors associated with smoking prevalence among Canadians. Health Rep 1992; 4:7–24. 6. Bearnall S, Edwards N. Social and cultural determinants of smoking behaviour in selected immigrant groups: Results of key informant interviews. Fam Commun Health 1995;18:65– 72. 7. Chen J, Ng E, Wilkins R. The health of Canada’s immigrants in 1994 –95. Health Rep 1996;7:33– 44. 8. East York Health Unit. Health Status Report. Toronto, Ontario, Canada, Toronto Public Health, East York Office, 1995. 9. Cheung M, Yau M, Zeigler S. The 1991 Every Secondary School Survey, Part 2: Detailed profiles of Toronto’s secondary school students. Toronto, Ontario, Canada, Toronto Board of Education, Research Services. 10. East York Health Unit. Highlights Report: High School Health Survey, Toronto, Ontario, Canada, 1996.
288
HOLOWATY ET AL.
11. Feldman L, Holowaty P, Harvey B, Shortt L. Alcohol use beliefs and behaviors among high school students. J Adolesc Health 1999;24:48 –58. 12. Feldman L, Holowaty P, Harvey B, et al. Physical activity beliefs and behaviors in high school students: Lifestyle factors related activity levels. Toronto Public Health, internal report, 1998. 13. Feldman L, Rannie K, Jamal A, et al. Sexual health and lifestyle behaviors in adolescent virgins and non-virgins. J Hum Sex 1997;6:197–210. 14. Dean AG, Dean JA, Coulombier D. EPI INFO Version 6: A Word Processing, Database, and statistics Program for Epidemiology on Computers: Atlanta, GA: Center for Disease Control and Prevention, 1994. 15. East York Health Unit. Report of the Tobacco Free High School Survey. Toronto Public Health Unit: East York Office, 1990. 16. Nova Scotia Council on Smoking Health. Smoke Free Nova Scotia Questionnaire for students of grades 6, 7, 8, 10 & 12. Halifax, Nova Scotia, Canada, Nova Scotia Council on Smoking and Health, 1994. 17. Social Evaluation Group. The Health of Canada’s Youth. Kingston, Ontario, Canada, Social Evaluation Group: Queen’s University, 1992. 18. Flay B, Koepke D, Thompson S, et al. Six year follow-up of the first Waterloo school smoking prevention trial. Am J Public Health 1989;79:1371– 6. 19. The Ontario Health Survey 1990 Highlights Report. Premier’s Council on Health, Well Being and Social Justice. Toronto, Ontario, Canada, Ontario Ministry of Health, September 1992. 20. Thunder Bay Health Department. The Thunder Bay Health Unit: Adolescent Smoking Survey. Thunder Bay, Ontario, Canada, 1993. 21. Offord D, Boyle M. Ontario Child Health Study: Children at Risk. Toronto, Ontario, Canada: Queen’s Printer for Ontario, 1989. 22. Norusse MJ. SPSS for Windows. Base Systems User’s Guide. Rel. 6.0. Chicago, IL: SPSS, 1993. 23. Stephens T. Youth smoking survey, 1994: Technical report. Ottawa, Ontario, Canada, Ministry of Supply and Services, 1996.
JOURNAL OF ADOLESCENT HEALTH Vol. 27, No. 4
24. Adlaf E. Student Drug and Alcohol Use 1977–1995. Toronto, Ontario, Canada: Addiction Research Foundation. 1996. 25. Patrick DL, Cheadle A, Thompson DC, et al. The validity of self-reported smoking: A review and meta-analysis. Am J Public Health 1994;84:1086 –93. 26. Barnea WR, Rahav G, Teichman M. The reliability and consistency of self-reports on substance use in a longitudinal study. Br J Addict 1987;82:891– 8. 27. U.S. Department of Health and Human Services. Preventing Tobacco Use Among Young People: A Report of the Surgeon General. Atlanta, Georgia, U.S. Department of Health and Human Services; Public Health Service, Centers for Disease Control and Prevention 1994;1–156. 28. Landrine H, Richardson JL, Klonoff EA, Flay B. Cultural diversity in the predictors of adolescent cigarette smoking: The relative influence of peers. J Behav Med 1994;17: 331– 46. 29. Bettes BA, Dusenbury L, Kerner J, et al. Ethnicity and psychosocial factors in alcohol and tobacco use in adolescents. Child Dev 1990;61:557– 65. 30. Juon HS, Shin Y, Nam JJ. Cigarette smoking among Korean adolescents: Prevalence and correlates. Adolescence 1995;30: 631– 42. 31. Li X, Fang X, Stanton B. Cigarette smoking among Chinese adolescents and its association with demographic characteristics, social activities, and problem behaviors. Subst Use Misuse 1996;31:545– 63. 32. de Moor C, Elder JP, Young RL, et al. Generic tobacco use among four ethnic groups in a school age population. J Drug Addict 1989;19:257–70. 33. Koepke D, Flay B, Johnson CA. Health behaviors in minority families: The case of cigarette smoking. Fam Commun Health 1990;13:35– 43. 34. Taylor KL, Kerner JF, Gold KF, Mandelblatt JS. Ever vs never smoking among an urban, multiethnic sample of Haitian-, Caribbean-, and US-born Blacks. Prev Med 1997; 26:855– 65. 35. Riley WT, Barenie JT, Mabe PA, Myers DR. The role of race and ethnic status on the psychological correlates of smokeless tobacco use in adolescent males. J Adolesc Health 1991;12:15–21.