Health Promotion International Advance Access published online on June 21, 2007
Health Promotion International, doi:10.1093/heapro/dam018
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Article |
Determinants of smoking among adolescents in the Southern Cape-Karoo region, South Africa
1 Department of Health Education and Health Promotion 2 Department of Experimental Psychology, Maastricht University, The Netherlands 3 Child, Youth, Family and Social Development, Human Sciences Research Council, Private Bag X07, Dalbridge 4014, South Africa 4 National Health Promotion Research and Development Group, Medical Research Council, South Africa 5 Department of Public Health and Clinical Medicine, Epidemiology and Public Health Sciences 6 Department of Clinical Sciences and Pediatrics, Umeå University, Sweden
* Corresponding author. E-mail: spanday{at}hsrc.ac.za
 |
SUMMARY |
|---|
 TOP SUMMARY INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Tobacco control programmes in multi-ethnic societies must take into account ethnic differences in the determinants of smoking. The I-Change Model, an extension of the Theory of Planned Behaviour, was used to investigate the factors related to smoking among a sample of 3378 Black African, Coloured and White, monthly and non-monthly smokers in the Southern Cape-Karoo Region, South Africa. Across the ethnic groups, non-monthly smokers reported a more positive attitude towards non-smoking, social influences that were more supportive of non-smoking, higher self-efficacy in stressful, routine and social situations, greater intention not to smoke in the next year and lower levels of depressive mood and risk behaviour. Regression analyses suggested that the weight of these determinants may differ in predicting monthly smoking among the ethnic groups. Black African students may benefit from the development of attitudinal cognitions and coping skills to counter peer influence. Coloured students also require skills to resist peer influence. White students require coping skills in stressful and social situations. Although there are more common than unique determinants of smoking among South African adolescents, further research is needed to understand the influence of differing social, economic and cultural contexts on smoking onset.
Key words: adolescence; smoking; determinants; ethnic differences
|
INTRODUCTION |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The rising levels of tobacco use among young people in the 1990s led to smoking being labelled a paediatric disease or epidemic (Perry et al., 1994
). Global estimates indicated that on average, 33% of adolescents had tried smoking and that 18.7% were past-month smokers (The Global Youth Tobacco Survey Collaborative Group, 2002). Similarly, studies in South Africa (SA) reported that 37.6% of adolescents had tried smoking in their lifetime and that 18.5% were past-month smokers (Swart et al., 2004).
Although there is some support for the long-term effectiveness of the social influences approach for smoking prevention programmes, lack of significant programme effects and variability in internal and external validity of several studies make this conclusion tenuous (Skara and Sussman, 2003). Psychological models addressing the cognitions of smoking, such as the Integrated Model of Change (the I-Change Model) (de Vries et al., 2003b), may be an effective way to develop smoking prevention programmes. The model incorporates insights from the Theory of Reasoned Action (Fishbein and Ajzen, 1975), Social Cognitive Theory (Bandura, 1986) and the Transtheoretical Model (Prochaska and DiClemente, 1983). The model has been used to assess the determinants of smoking and to develop and evaluate smoking prevention programmes in several European countries (de Vries et al., 1994; de Vries et al., 1995; de Vries et al., 1988; Dijkstra et al., 1999). The model assumes that the most important determinant of behaviour is behavioural intention, which, in turn, is influenced by three proximal factors, namely, attitudes, social influences and self-efficacy. Distal factors such as demographic (e.g. age, gender) and psychological factors (e.g. depression) are assumed to influence behaviour via the proximal factors.
SA requires data on the psychosocial determinants of smoking for the development of smoking prevention programmes. Demographic factors such as age at initiation, gender and school performance; cognitive factors such as attitude towards smoking, social influences such as parental and peer smoking; self-efficacy and intention to smoke have been linked to adolescent smoking onset and smoking progression [see, e.g., reviews by Conrad et al. (1992) and Tyas and Pederson (1998)]. Additionally, psychological factors such as depression (Escobedo et al., 1998; Horn et al., 2004) and risk behaviours such as alcohol and marijuana use (US Department of Health and Human Services, 1994; Holm et al., 2003; Siqueira and Brook, 2003; Wetzels et al., 2003) have been shown to co-vary with tobacco use.
Ethnic differences have also been reported in the prevalence and determinants of smoking (US Department of Health and Human Services, 1994; Griesler and Kandel, 1998; Kelder et al., 2003; Kandel et al., 2004). Lower smoking rates have been consistently reported among Black African1 adolescents (15.7%) in SA compared with White (21.7%) and Coloured (38.7%) adolescents (Swart et al., 2003; Swart et al., 2004). However, limited data are available on the mechanisms underpinning these differences and whether programmes developed primarily for White adolescents are applicable to other ethnic groups (US Department of Health and Human Services, 1998; Skara and Sussman, 2003).
The history of Apartheid in SA means that poverty and inequality continue to exhibit strong spatial and racial biases (United Nations Population Development South Africa, 2003). Consequently, ethnicity has become a proxy for social, economic, spatial and cultural differences when these factors are difficult to estimate. In the present study, ethnic differences in the motivational determinants of smoking onset were investigated with a view to develop culturally sensitive tobacco control programmes for adolescents (US Department of Health and Human Services, 1998; Skara and Sussman, 2003; Markam et al., 2004; Swart et al., 2004).
|
METHODS |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Participants and sampling
In 2002, a cross-sectional survey was administered to grade 9–11 students (14–16 years of age) in the Southern Cape-Karoo Region, Western Cape Province. Forty-two public schools were eligible to participate in the study. On the basis of previous research (Panday et al., 2003
), school selection was stratified by ethnicity in accordance with the school's previous race classification, namely, Black African (six schools), Coloured (17 schools) or White (19 schools). Previous research (Swart et al., 2003) was also used to estimate the sample of 100 smokers required for each ethnic by gender group. A total of 23 schools were selected to participate in the study, consisting of all six Black African schools, eight randomly selected Coloured schools and nine randomly selected White schools. The number of classes selected was proportional to the number of grade 9–11 classes in the school, on the basis of an estimated class size of 40 students. All students in the selected classes were eligible to participate in the study. A total of 121 classes representing 4768 students were selected to participate in the study.
The Research Ethics Committees of the South African Medical Association and the Medical Faculty of Umeå University granted ethical approval for the study. Permission was also obtained from the education department, principals of the selected schools and from parents and students in the selected classes. Parents and students were informed both verbally and in writing that all answers would be treated confidentially and viewed by the researchers only.
Questionnaire
The questionnaire was administered in three languages, during two regular classroom periods either to individual classes or to groups of classes. The questionnaire was prepared in English and translated to Afrikaans and to Xhosa. To ensure the accuracy of the translations, the Afrikaans and Xhosa versions of the questionnaire were back translated to English by a second first-language speaker. Discrepancies were discussed and consensus reached. To guarantee confidentiality, trained research assistants administered the survey in the classroom. Teachers were asked to leave the classroom during survey administration. Students were requested not to write their names on the questionnaires to ensure their anonymity.
The questionnaire was based on the European Smoking Prevention Framework Approach (ESFA) questionnaire (de Vries et al., 2003a; de Vries et al., 2003b). The questionnaire assessed smoking behaviour, attitude, social influences, self-efficacy expectations, intention not to smoke in the next year, depressive mood, risk behaviour and several demographic items. Item selection and formulation were validated and localized to the South African context through data obtained from prior qualitative research (Panday et al., 2003), through focus group discussions conducted during questionnaire development and from a pilot study conducted among 292 grade 9 students from the three language groups.
Measures
Smoking behaviour
Adolescents were asked to pick a statement that best described them, out of a set of specific smoking-related questions. They were then categorized as monthly smokers (smoking at least once a month or more) or non-monthly smokers (never smoked, not even one puff, try smoking once in a while and quit smoking) (Kremers et al., 2004). Self-reports of smoking could not be biologically validated owing to logistical and financial constraints. However, when anonymity is assured, self-reports have been shown to be reliable and in agreement with biochemical markers (Dolcini et al., 1996). Self-reported smoking was cross-validated using an algorithm consisting of four additional concepts measuring current smoking and lifetime smoking.
Attitudes
Their attitudes were assessed by two five-point scales measuring the pros and cons of non-smoking, which were identified using factor analyses (alpha factoring and direct oblimin rotation) and through previous research (Kremers et al., 2001a; Kremers et al., 2001b; de Vries et al., 2003a). The pros of non-smoking were measured with a seven-item scale [e.g. ‘If I do not smoke it will be very good for my health (+2) or very bad for my health (–2)’; Cronbach's alpha (
) = 0.86]. The cons of non-smoking were measured with a five-item scale [e.g. ‘If I do not smoke it will make me feel very relaxed (–2) or very stressed (+2)’; = 0.68]. One item referring to drug use was excluded owing to ambiguity in the Xhosa version of the questionnaire.
Social influences
Influences from the social environment were measured by assessing social norms, social modelling and social pressure from important others: mother, father, grandmother, grandfather, brother(s), sister(s), friends, best friend, classmates and teachers. Social norms were measured on a five-point scale assessing students' perceptions of whether important others (combined into family, six items, = 0.86; friends, three items, = 0.79; teacher) thought that they definitely should (–2) or should not (+2) smoke. Social modelling assessed the perceived smoking behaviour of mother, father, grandmother, grandfather, brother(s), sister(s) and best friends, using a two-point scale (0, smoking; 1, non-smoking). The seven items were analysed separately, as they did not load uniquely on the social modelling factor. Social pressure was measured on a five-point scale to assess how often social pressure not to smoke (0, never; 4, very often) was encountered (combined into family, six items, = 0.90; friends, three items, = 0.87; teacher).
Self-efficacy
Multiple-item scales measured how sure (+2, sure that I will not smoke; –2, sure that I will smoke) students felt that they could refrain from smoking in stressful situations (stress self-efficacy, 10 items, = 0.98), routine situations (routine self-efficacy, five items, = 0.94) and social situations (social self-efficacy, four items, = 0.93). The items were identified through previous research (de Vries et al., 1988; Lawrance, 1998).
Demographic variables
Characteristics of the participants were provided by asking for ethnicity (1, Black African; 2, Coloured; 3, White), gender (0, boys; 1, girls), age (continuous scores), school performance (0, lowest; 1, average; 2, best), spending money (0, no; 1, yes) and religion (0, no religion; 1, religious).
Depressive mood
Kandel and Davies' (Kandel and Davies, 1982) scale was used to measure depressive mood. Six items using a five-point scale (0, never; 4, always) assessed how often adolescents were bothered or troubled by the following states, namely ‘feeling too tired to do things’, ‘having trouble going to sleep or staying asleep’, ‘feeling unhappy, sad or depressed’, ‘feeling hopeless about the future’, ‘feeling nervous or tense’ and ‘worrying too much about things’. Mean scores were calculated to produce an index of depressive mood with a range of 0–4 ( = 0.87).
Risk behaviour
Risk behaviour was measured with eight items on a five-point scale (0, never; 1, sometimes; 2, less than once a month; 3, not weekly but at least once a month; 4, at least once a week) assessing use of alcohol, marijuana, methaqualone, other drugs and sniffing substances, gambling, playing the Lotto (form of lottery in SA) and playing truant from school; = 0.69.
Intention
Participants' intention not to smoke in the next year was measured by one item on a five-point scale (–2, definitely yes; +2, definitely not).
Statistical analyses
For each ethnic group, monthly smokers (coded as 1) and non-monthly smokers (coded as 0) were compared with regard to demographic variables, using logistic regression for dichotomous variables and F-tests for continuous variables. Significant differences between monthly and non-monthly smokers were found for age, gender, school performance, spending money and religion. These variables were included as covariates in subsequent analyses. Differences between the smoking categories and ethnic groups on the psychosocial variables were analysed using 2 (smoking categories: non-monthly smokers versus monthly smokers) x 3 (ethnic groups: Black African versus Coloured versus White) covariance analyses (ANCOVAs). Where mean scores were dependent on the interaction of smoking categories and ethnic groups, simple main effect analyses were conducted to test the relationship between smoking categories and the dependent variable within each ethnic group. Where a significant main effect for ethnic group was found (corrected for smoking status), contrast analyses were conducted to determine which ethnic groups differed significantly.
Owing to several interaction effects between attitude, social influences and self-efficacy and ethnicity, separate logistic regression analyses were run for each ethnic group to identify the correlates of monthly smoking using smoking status as the dependent variable. The high intercorrelations among the subscales of social norms, social pressure and self-efficacy resulted in the subscales being combined into one scale each. Social pressure was excluded from the regression analyses owing to a suppressor effect among Black African students. A backward deletion procedure was used to determine the final model of variables that relate to smoking status for each ethnic group. In accordance with the theoretical model used in this study, demographic variables as well as distal factors (depressive mood and risk behaviour) were entered in Block 1, attitude, social influences and self-efficacy variables in Block 2 and intention not to smoke in the next year in Block 3 (Holm et al., 2003). Owing to the large sample size, the significance level was set at p < 0.01.
|
RESULTS |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Characteristics of the sample
Of the 4768 students selected to participate in the study, 3869 completed the questionnaire. Owing to missing or incomplete data on key variables, 491 cases were excluded from the analyses, leaving 3378 (559 monthly smokers and 2990 non-monthly smokers) eligible for the present study. The mean age of the participants was 16.23 years (SD = 1.65) and the overall distribution of males and females was 58.2% for males and 41.8% for females. Most students were Black African (52.3%, n = 1857), followed by Coloured (38.3%, n = 1358) and White (9.4%, n = 334) students. Most students reported that they received spending money (81.0%), that they were religious (80.5%) and had an average school performance (58.5%).
Differences between monthly and non-monthly smokers
Across the ethnic groups, non-monthly smokers reported more pros of non-smoking, fewer cons of non-smoking, more social norms not to smoke from family, friends and teachers, more social pressure to refrain from smoking from friends and teachers, fewer important others who smoked, greater self-efficacy during stressful, routine and social situations, a more positive intention not to smoke in the next year and lower levels of depressive mood and risk behaviour (Table 1).
|
Within each ethnic group, monthly and non-monthly smokers differed on the pros of non-smoking, stress and social self-efficacy scales, intention not to smoke in the next year, depressive mood and risk behaviour. Simple main effect analyses showed that for all ethnic groups, non-monthly smokers were more convinced of the pros of non-smoking than monthly smokers, displayed higher stress and social self-efficacy and had a more positive intention not to smoke in the next year. However, these patterns were more pronounced among Coloured and White students than among Black African students.
Additionally, Black African and Coloured monthly smokers reported higher scores on the depressive mood scale than Black African and Coloured non-monthly smokers. For all three ethnic groups, monthly smokers engaged in risk behaviour to a greater extent than non-monthly smokers, although the difference was more pronounced among Black African students than among Coloured and White students.
Differences between ethnic groups
Independent of smoking status, Black African students were less convinced of the pros of non-smoking than Coloured and White students, whereas Coloured students were least convinced of the cons of non-smoking, followed by Black African and, in turn, White students (Table 2).
|
Coloured students experienced a stronger social norm not to smoke from their friends than Black African and White students, whereas Coloured and White students experienced stronger norms not to smoke from their family and teachers than Black African students. Coloured students also reported more smoking among their important others than Black African and White students. Most Coloured students, followed by Black African students, and, in turn, White students experienced pressure from their family and friends not to smoke. More Coloured students than Black African and White students experienced pressure from their teacher not to smoke.
White and Coloured students reported lower stress and social self-efficacy than Black African students, whereas Coloured students reported lower routine self-efficacy than the other two groups. Black African students expressed a more positive intention not to smoke in the next year than Coloured and White students. White students reported the highest level of depressive mood, followed by Coloured and, in turn, Black African students. The ethnic groups did not differ significantly on their engagement in risk behaviours.
To determine the most important factors associated with monthly smoking, separate logistic regression analyses were conducted for each ethnic group (Table 3).
|
Among the Black African students, in the first step, gender, depressive mood and risk behaviour were significant correlates. In the second step, gender and depressive mood remained as significant correlates, as were having a non-smoking best friend and self-efficacy. In the third step, gender, depressive mood, a non-smoking best friend and self-efficacy were significant correlates. The model correctly classified 47.9% of the monthly smokers and 98.4% of the non-monthly smokers.
Among the Coloured students, in the first step, depressive mood and risk behaviour were significant correlates of monthly smoking. In the second step, risk behaviour, a non-smoking best friend and self-efficacy were significant correlates of monthly smoking. In the third step, risk behaviour, a non-smoking best friend, self-efficacy and intention not to smoke were significant correlates of monthly smoking. The model correctly classified 71.9% of the monthly smokers and 95.0% of the non-monthly smokers.
Among the White students, in the first step, risk behaviour was the only significant correlate of monthly smoking. In the second step, risk behaviour and self-efficacy emerged as significant correlates of monthly smoking. In the third step, risk behaviour, self-efficacy and intention not to smoke in the next year were significant correlates of monthly smoking. The model correctly classified 77.1% of monthly smokers and 95.0% of non-monthly smokers.
|
DISCUSSION |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This study describes the differences between Black African, Coloured and White, monthly and non-monthly smokers on the psychosocial determinants of smoking. Overall, non-monthly smokers displayed a more positive attitude towards non-smoking, were surrounded by a social environment that was supportive of non-smoking and displayed higher self-efficacy not to smoke in stressful, routine and social situations. Non-monthly smokers were also more positive about their intention not to smoke in the next year and reported lower levels of depressive mood and risk behaviour. Holm et al. (Holm et al., 2003
) reported similar findings among Danish smokers and non-smokers.
The interaction patterns between smoking status and the ethnic groups on the pros of non-smoking, self-efficacy subfactors and intention not to smoke in the next year suggest that smoking status had a different relationship with these factors in each ethnic group. The differences between monthly and non-monthly smokers on these factors were more pronounced among Coloured and White students than among Black African students. For the other cognitive factors, the lack of interaction patterns between smoking status and the ethnic groups suggest that the differences between monthly and non-monthly smokers were similar within each ethnic group. Other studies have also concluded that there are more common than unique ethnic predictors of smoking initiation (Griesler et al., 2002; Kandel et al., 2004).
Black African students, however, reported lower scores on the pros of non-smoking, suggesting that they were on average less convinced of the detrimental effects of smoking. The lack of intention as a correlate of smoking and the weak fit of the research model in this group also intimate towards distal correlates of monthly smoking. As predicted by the I-Change Model, for Coloured students, the distal correlate (depressive mood) influenced monthly smoking through the psychosocial variables. However, for Black African students, the distal correlates (gender and depressive mood) had a direct influence on monthly smoking. Although the I-Change Model provides a useful framework for the development of tobacco control programmes for Coloured and White students, further investigation is required of the attitudinal beliefs, cultural as well as distal factors motivating smoking among Black African adolescents. It is possible that the belief components of smoking onset for this group were not adequately measured by the psychosocial determinants included in the study.
Even though self-efficacy was a correlate of smoking for all ethnic groups, the covariance analyses showed that Coloured students may require coping skills for all of the self-efficacy subfactors while White students may benefit mostly from these skills in stressful and social situations. The covariance and regression analyses also highlight the importance of the social environment for Coloured and Black African students. Coloured students reported higher perceived social norms and social pressure not to smoke, accompanied by higher levels of smoking among their important others. A smoking best friend was a correlate of monthly smoking for both Coloured and Black African students but not for White students. Even though studies have reported peer smoking as predictive of smoking for White students (Sussman et al., 1987; Griesler and Kandel, 1998; Unger et al., 2001), qualitative research in SA demonstrated that White smokers, unlike Black African and Coloured smokers, believed that they smoked by choice and were not influenced by their peers (Panday et al., 2003). Community-based programmes in SA must account for the differing influence of the social environment on smoking onset. In fact, social norms and modelling can account for the influence of ethnicity on adolescent smoking (Ellickson et al., 2003).
For Coloured and White students, risk behaviour was the only distal correlate that was not related to monthly smoking through the psychosocial variables. Danish adolescents also reported an association between smoking and deviant behaviour (Holm et al., 2003). The co-variation and gateway effect of smoking and risk behaviour are well documented (US Department of Health and Human Services, 1994; Everett et al., 1998; Rigotti et al., 2000; Chen et al., 2002; Siqueira and Brook, 2003; Wetzels et al., 2003). However, further investigation is warranted in SA to inform the need for comprehensive rather than singular programmes.
Smoking prevalence has migrated over time through the socio-economic groups and resulted in higher smoking rates among those from lower socio-economic groups (World Bank, 1999). Factors, such as low socio-economic status, poverty and low educational attainment, that are related to smoking also tend to cluster in certain ethnic groups (Tyas and Pederson, 1998). However, the impact of poverty, racial discrimination, lack of opportunity and low perceived life chances on adolescent risk behaviour has not been explored fully (Jessor, 1991). As this study could not include robust measures of socio-economic status and poverty, because they are difficult to estimate among adolescents, ethnicity was used as a proxy for the set of factors. Even though the findings showed that the relative weight of the motivational factors may be different for the ethnic groups, the history of racial segregation and discrimination in SA makes recommendations for ethnic-specific school-based programmes undesirable. Thus, the direct translation of the ethnic factors related to smoking onset into programme objectives would be an over-simplification of the determinants of ethnic differences. However, further understanding of the link between other factors, such as low socio-economic status, poverty and low educational attainment, that tend to aggregate in certain ethnic groups will allow for tobacco control programmes to be tailored to the motivational determinants of these subgroups. In addition, the cross-sectional nature of the study cannot estimate causality and must be replicated in longitudinal studies.
|
ACKNOWLEDGEMENTS |
|---|
The authors would like to thank the participating schools and the students as well as the research assistants for assisting with data collection. We are grateful to the Medical Research Council, South Africa, for funding this study.
|
FOOTNOTES |
|---|
1 During the Apartheid years, all South Africans were classified into race groups in accordance with the Population Registration Act of 1950, namely, Black African (people of African descent), Coloured (people of mixed descent), Indian (people of Indian descent) and White (people of European descent). The authors in no way subscribe to this classification.
|
REFERENCES |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Bandura A. Social Foundations of Thought and Action: A Social Cognitive Theory (1986) New York: Prentice Hall.
Chen X., Unger J. B., Palmer P., Weiner M. D., Johnson C. A., Wong M. M., et al. Prior cigarette smoking initiation predicting current alcohol use: evidence for a gateway drug effect among California adolescents from eleven ethnic groups. Addictive Behaviors (2002) 27:799–817.[CrossRef][Web of Science][Medline]
Conrad K. M., Flay B. R., Hill D. Why children start smoking cigarettes: predictors of onset. British Journal of Addiction (1992) 87:1711–1724.[CrossRef][Web of Science][Medline]
de Vries H., Dijkstra M., Kuhlman P. Self-efficacy: the third factor besides attitude and subjective norm as a predictor of behavioural intentions. Health Education Research (1988) 3:273–282.
de Vries H., Backbier E., Dijkstra M., Van Breukelen G., Parcel G., Kok G. A Dutch social influence smoking prevention approach for vocational school students. Health Education Research (1994) 9:365–374.
de Vries H., Backbier E., Kok G. The impact of social influences in the context of attitude, self-efficacy, intention and previous behavior as predictors of smoking onset. Journal of Applied Social Psychology (1995) 25:237–257.[Medline]
de Vries H., Mudde A., Kremers S., Wetzels J., Uiters E., Ariza C., et al. The European Smoking Prevention Framework Approach (ESFA): short-term effects. Health Education Research (2003a) 18:649–663.
de Vries H., Mudde A., Leijs I., Charlton A., Vartiainen E., Buijs G., et al. The European Smoking Prevention Framework Approach (EFSA): an example of integral prevention. Health Education Research (2003b) 18:611–626.
Dijkstra M., Mesters I., de Vries H., Van Breukelen G., Parcel G. S. Effectiveness of a social influence approach and boosters to smoking prevention. Health Education Research (1999) 14:791–802.
Dolcini M. M., Adler N. E., Ginsberg D. Factors influencing agreement between self-reports and biological measures of smoking among adolescents. Journal of Research on Adolescence (1996) 6:515–542.[Web of Science]
Ellickson P. L., Perlman M., Klein D. J. Explaining racial/ethnic differences in smoking during the transition to adulthood. Addictive Behaviors (2003) 28:915–931.[CrossRef][Web of Science][Medline]
Escobedo L. G., Reddy M., Giovino G. A. The relationship between depressive symptoms and cigarette smoking in US adolescents. Addiction (1998) 93:433–440.[CrossRef][Web of Science][Medline]
Everett S. A., Giovino G. A., Warren C. W., Crossett L., Kann L. Other substance use among high school students who use tobacco. Journal of Adolescent Health (1998) 23:289–296.[CrossRef][Web of Science][Medline]
Fishbein M., Ajzen I. Belief, Attitude, Intention and Behavior: An Introduction to Theory and Research (1975) MA: Addison Wesley, Reading.
Griesler P. C., Kandel D. B. Ethnic differences in correlates of adolescent cigarette smoking. Journal of Adolescent Health (1998) 23:167–180.[CrossRef][Web of Science][Medline]
Griesler P. C., Kandel D. B., Davies M. Ethnic differences in predictors of initiation and persistence of adolescent cigarette smoking in the National Longitudinal Survey of Youth. Nicotine and Tobacco Research (2002) 4:79–93.[CrossRef]
Holm K., Kremers S. P., de Vries H. Why do Danish adolescents take up smoking? European Journal of Public Health (2003) 13:67–74.
Horn K., Dino G., Kalsekar I., Massey C. J., Manzo Tennant K., McGloin T. Exploring the relationship between mental health and smoking cessation: a study of rural teens. Prevention Science (2004) 5:113–126.[CrossRef][Web of Science][Medline]
Jessor R. Risk behavior in adolescence: a psychosocial framework for understanding and action. Journal of Adolescent Health (1991) 12:597–605.[CrossRef][Web of Science][Medline]
Kandel D. B., Davies M. Epidemiology of depressive mood in adolescents. Archives of Pediatrics and Adolescent Medicine (1982) 39:1205–1212.
Kandel D. B., Kiros G. E., Schaffran C., Hu M. C. Racial/ethnic differences in cigarette smoking initiation and progression to daily smoking: a multilevel analysis. American Journal of Public Health (2004) 94:128–135.
Kelder S. H., Prokhorov A., Barroso C. S., Murray N., Orpinas P., McCormick L. Smoking differences among African American, Hispanic, and White middle school students in an urban setting. Addictive Behaviors (2003) 28:513–522.[CrossRef][Web of Science][Medline]
Kremers S. P., Mudde A. N., de Vries H. ‘Kicking the initiation’: do adolescent ex-smokers differ from other groups within the initiation continuum? Preventive Medicine (2001a) 33:392–401.[CrossRef][Web of Science][Medline]
Kremers S. P., Mudde A. N., De Vries H. Subtypes within the precontemplation stage of adolescent smoking acquisition. Addictive Behaviors (2001b) 26:237–251.[CrossRef][Web of Science][Medline]
Kremers K., Mudde A., de Vries H. Development and longitudinal test of an instrument to measure behavioral stages of smoking initiation. Substance Use and Misuse (2004) 39:225–252.[CrossRef]
Lawrance L. Validation of a self-efficacy scale to predict adolescent smoking. Health Education Research (1998) 4:351–360.[CrossRef]
Markam W. A., Aveyard P., Thomas H., Charlton A., Lopez M. L., de Vries H. What determines future smoking intentions of 12- to 13-year-old UK African-Caribbean, Indian, Pakistani and white young people? Health Education Research (2004) 19:15–28.
Panday S., Reddy S. P., Bergstrom E. A qualitative study on the determinants of smoking behaviour among adolescents in South Africa. Scandinavian Journal of Public Health (2003) 31:204–210.[CrossRef][Web of Science][Medline]
Perry C. L., Eriksen M. P., Giovino G. Tobacco use: a pediatric epidemic. Tobacco Control (1994) 3:97–98.
Prochaska J. O., DiClemente C. C. Stages and processes of self-change of smoking: toward an integrative model of change. Journal of Consulting and Clinical Psychology (1983) 51:390–395.[CrossRef][Web of Science][Medline]
Rigotti N. A., Lee J. E., Wechsler H. US college students' use of tobacco products: results of a national survey. Journal of the American Medical Association (2000) 284:699–705.
Siqueira L. M., Brook J. S. Tobacco use as a predictor of illicit drug use and drug-related problems in Colombian youth. Journal of Adolescent Health (2003) 32:50–57.[CrossRef][Web of Science][Medline]
Skara S., Sussman S. A review of 25 long-term adolescent tobacco and other drug use prevention program evaluations. Preventive Medicine (2003) 37:451–474.[CrossRef][Web of Science][Medline]
Sussman S., Dent C. W., Flay B. R., Hansen W. B., Johnson C. A. Psychosocial predictors of cigarette smoking onset by white, Black, Hispanic, and Asian adolescents in Southern California. Morbidity and Mortality Weekly Report (1987) 36(Suppl. 4):11s–16s.
Swart D., Reddy P., Ruiter R. A., de Vries H. Cigarette use among male and female grade 8–10 students of different ethnicity in South African schools. Tobacco Control (2003) 12. E1.
Swart D., Reddy P., Panday S., Philip J., Naidoo N., Ngobeni N. The 2002 Global Youth Tobacco Survey (GYTS): The 2nd GYTS in South Africa (SA)—A Comparison between GYTS (SA) 1999 and GYTS (SA) 2002. (2004) Cape Town: Medical Research Council.
The Global Youth Tobacco Survey Collaborative Group. Tobacco use among youth: a cross-country comparison. Tobacco Control (2002) 11:252–270.
Tyas S. L., Pederson L. L. Psychosocial factors related to adolescent smoking: a critical review of the literature. Tobacco Control (1998) 7:409–420.
Unger J. B., Rohrbach L. A., Cruz T. B., Baezconde Garbanati L., Howard K. A., Palmer P. H., et al. Ethnic variation in peer influences on adolescent smoking. Nicotine and Tobacco Research (2001) 3:167–176.[CrossRef]
United Nations Population Development South Africa. South Africa Human Development Report 2003. The Challenge of Sustainable Development in South Africa: Unlocking People's Creativity. (2003) Cape Town: Oxford University Press.
US Department of Health and Human Services. Preventing Tobacco Use among Young People: A Report of the Surgeon General (1994) Atlanta. Author.
US Department of Health and Human Services. Tobacco Use among US Racial/Ethnic Minority Groups. A Report of the Surgeon General (1998) Atlanta. Author.
Wetzels J. J., Kremers S. P., Vitoria P. D., de Vries H. The alcohol–tobacco relationship: a prospective study among adolescents in six European countries. Addiction (2003) 98:1755–1763.[CrossRef][Web of Science][Medline]
World Bank. Curbing the Epidemic. Governments and the Economies of Tobacco Control (1999) Washington, DC: World Bank.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||