Asia-Pacific Journal of Public Health http://aph.sagepub.com/ Social Normative Beliefs About Smoking Among Vietnamese Adolescents Randy M Page, Nguyen Thanh Huong, Hoang Khanh Chi and Truong Quang Tien Asia Pac J Public Health 2012 24: 68 originally published online 24 May 2010 DOI: 10.1177/1010539510370993 The online version of this article can be found at: http://aph.sagepub.com/content/24/1/68 Published by: http://www.sagepublications.com On behalf of: Asia-Pacific Academic Consortium for Public Health Additional services and information for Asia-Pacific Journal of Public Health can be found at: Email Alerts: http://aph.sagepub.com/cgi/alerts Subscriptions: http://aph.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav >> Version of Record - Feb 16, 2012 OnlineFirst Version of Record - May 24, 2010 What is This? Downloaded from aph.sagepub.com by guest on October 11, 2013 Social Normative Beliefs About Smoking Among Vietnamese Adolescents Asia-Pacific Journal of Public Health 24(1) 68–81 © 2012 APJPH Reprints and permission: http://www sagepub.com/journalsPermissions.nav DOI: 10.1177/1010539510370993 http://aph.sagepub.com Randy M Page, PhD1, Nguyen Thanh Huong, MSc, MPH, PhD2, Hoang Khanh Chi, MPH2, and Truong Quang Tien, MD, MPH2 Abstract Tobacco-related deaths in Vietnam are forecast to climb from 40 000 annually to 70 000 by 2030 Previous research in Western nations has found social factors to be important determinants of adolescent smoking Because these factors remain unexplored in Vietnamese youth, the purpose of this study was to examine social normative beliefs regarding smoking in a school-based sample of North Vietnamese adolescents and the association of these factors with smoking behavior and susceptibility to smoking Three measures of normative beliefs regarding smoking were evaluated in cross-sectional surveys of secondary students Of the measures, parent/peer disapproval was the most consistent normative belief associated with smoking behavior and susceptibility to smoking Youth smoking prevention programs should consider assessing and taking into account normative beliefs and develop strategies that provide accurate information about the actual prevalence of smoking, the types of individuals who smoke, and approval/disapproval of smoking by parents and peers Keywords cigarette smoking, tobacco use, adolescents, normative beliefs, Vietnam Introduction Youth smoking in Vietnam and other Southeast Asian countries is generally reported in international studies to be lower than among youth in other regions of the world.1 The Global Youth Tobacco Survey in Vietnam found that 4.9% of students aged 13 to 15 years smoked cigarettes in the past month.2 Smoking among Vietnamese girls is quite rare with only 1.5% of 13- to 15-year-old girls reporting smoking in the past month, compared with 9.1% of 13- to 15-year-old boys.2 A contrast in smoking rates between males and females persists into adulthood and the rate of smoking among males escalates much higher than in adolescence in males, but remains low among adult females The National Health Survey3 in 2002 found that 56% of adult men and Brigham Young University, Provo, UT, USA Hanoi School of Public Health, Hanoi, Vietnam Corresponding Author: Randy M Page, Department of Health Science, Brigham Young University, 221 Richards Building, Provo, UT 84602, USA Email: randy_page@byu.edu Page et al 69 1.8% of adult women smoke regularly and the Vietnam Living Standard Survey4 found 50% of Vietnamese men and 3.5% of Vietnamese women smoked Smoking rates among Vietnamese male university students are also markedly higher than among female students, as well as among male medical students compared with female medical students.5 Public health concern about youth smoking is high in Vietnam More than 40 000 Vietnamese die each year from smoking; a result of the fact that Vietnam has one of highest rates of smoking prevalence among adult men in the world.6 According to the World Health Organization, despite Vietnam’s complete ban on tobacco advertising, tobacco-related deaths are forecast to climb to 70 000 by 2030 largely because of a failure to enforce the ban and shifting promotional tactics used by the tobacco industry to target youth in tobacco product marketing.7 Thus, the high burden of disease and death from lung cancer and other tobacco-induced disease in Vietnam is likely to continue unless there is a significant reduction in smoking in both adult and youth populations, and vigilant efforts need to be undertaken to prevent smoking in young Vietnamese women so that smoking rates not climb Considering the need for tobacco use prevention in Vietnam, it is disheartening that there is essentially no research reported in the international literature investigating determinants of smoking among Vietnamese youth Research conducted primarily in Western countries has found a range of factors to be associated with the onset and maintenance of youth smoking, and social factors have been found to be among the most important determinants of adolescent smoking.8 Thus, research examining youth smoking should not overlook the social nature of risk behaviors in adolescence.9 Peer effects have been found to be potent determinants of adolescent smoking10,11 and there is greater risk of substance use when a young person is rooted in a peer network where substance use is normative.12 Cigarette smoking often becomes a habit in the company of peers who smoke.13,14 Adolescents often evaluate their personal behavior in comparison with their perception of the behaviors of peers This process of social comparison helps them to adjust to peer norms and behaviors.15 According to Piko,16 adolescents’ estimates of their peers’ smoking prevalence can influence their own smoking behavior As youth are motivated to conform to peer norms, this tendency to overestimate the prevalence of peers’ smoking represents a high risk factor for adolescents.17,18 The current study investigates social normative beliefs regarding smoking in a sample of youth in the north of Vietnam and the association of these factors with smoking behavior and susceptibility to smoking Specifically, this study follows a methodological approach used by Primack et al19 to measure social normative beliefs based on health behavior theory and models,20 which incorporates measures of smoking normative beliefs theorized to be independently associated with adolescent smoking behavior and susceptibility to future smoking These measures of normative beliefs are perceived prevalence of smoking, perceived popularity of smoking among successful/elite elements of society, and perceived disapproval by friends and family Subjective normative beliefs are important explanatory variables of smoking behavior in the theory of planned behavior,21 problem behavior theory,22 and social cognitive theories,23 where the acceptance of smoking by significant others has been shown to predict smoking among various groups of adolescents.24,25 As a result, it was hypothesized that these measures would be independently associated with smoking and susceptibility to smoking in this sample of adolescents Methods Sample The total sample consisted of 2000 students in grades 10 to 12 in high schools The mean age for the total sample was 16.8 years (SD = 1.2), with a range of 15 to 19 years, and 858 were male, 70 Asia-Pacific Journal of Public Health 24(1) 1005 were female, and 137 did not report their gender In the total sample, there were 604 10th graders, 667 11th graders, and 728 12th graders (1 did not report the grade level) One school is located in Hai Duong Province, in the Thanh Mien District and consists of 1009 students in these grades in attendance on the day that the survey was administered The survey was administered to all students enrolled in the school, with the exception of those who were absent during survey administration The Thanh Mien subjects included 403 males (43.6%) and 521 females (56.4%) and 85 not reporting their gender The grade level distribution of the subjects was 29.9% (302) 10th graders, 27.2% (274) 11th graders, and 42.9% (433) 12th graders The mean age of Thanh Mien students was 16.9 years (SD = 1.2) Hai Duong Province is a rural, agricultural province in the northern river delta and is located 58 kilometers from Hanoi The province has a population of 1.8 million people, and the Thanh Mien District has a population of 131 000 The other school included in the sample is in Hanoi, which has a population estimated to exceed million This school is located in the Ba Dinh District, which is an urban, densely populated district and includes 991 students who were in 30 randomly selected classes, representing half of the classes in the school The 991 students represent approximately 47% of the total school population of 2110 students This sampling procedure was used in order to obtain an equivalent number of subjects from the Hanoi school as the Hai Duong school In the sample were 455 males (48.5%) and 484 females (51.5%), with 53 subjects not reporting their gender The grade level distribution of the Hanoi subjects was 30.5% (302) 10th graders, 39.7% 11th graders (393), and 29.8% 12th graders (295), with subject not reporting grade level The mean age of Hanoi students was 16.7 years (SD = 1.2) The mean age of the students in the schools did not differ but there was a higher percentage of female students in the Thanh Mien school Thanh Mien had a higher percentage of 12th grade students, and the Hanoi school had a higher percentage of 11th grade students The proportion of 10th grade students in the schools was nearly identical (29.9% and 30.5%) Survey Administration Permission to conduct this study was obtained from school officials at the schools The survey was administered in classes to students in grades 10 to 12 by personnel from the Hanoi School of Public Health Students were informed that their participation in the survey was voluntary and that they could refuse to participate if they desire There were no reports of students refusing to participate in the survey Students were also informed not to put their name on the survey questionnaire and to place their completed survey forms in an envelope, which was subsequently sealed by the university personnel administering the survey This ensured that completed surveys were not seen by teachers or other school personnel Teachers were not present in the class during the questionnaire administration Survey administration and other study procedures were approved by the Institutional Review Board for Human Subjects at Brigham Young University Instrumentation The measures of smoking normative beliefs were obtained from Primack et al19 and formed the basis for the survey instrument Primack et al.19 describe that these normative belief measures were developed after a comprehensive literature review and based on input from experts in tobacco control, public health, and adolescent medicine and also focus groups with adolescents Three of the items measured perceived disapproval from family and friends and used a 4-point Likert-type scale (strongly agree, agree, disagree, and strongly disagree) for the following: (a) according to my parents, it is very important for me to not smoke cigarettes; (b) according to my friends, it is very important for me to not smoke cigarettes; and (c) according to most people my age, it is very Page et al 71 important for me to not smoke cigarettes These perceived disapproval items were based on items from the Fishbein–Ajzen–Hansen Questionnaire developed by the authors of the theory of planned behavior.21,26 The perceived disapproval from family and friends scale was formed by summing the three items and taking the mean as the score The internal consistency score (Cronbach’s a) was 85 Four items measured perceived popularity among successful/elite elements of society and were also constructed on a 4-point Likert-type scale (strongly agree, agree, disagree, and strongly disagree) and included the following: (a) most successful businesspeople smoke cigarettes at least once a month; (b) in general, more “cool” people smoke cigarettes than “uncool” people; (c) wealthy people are more likely to smoke cigarettes than poor people; and (d) my favorite celebrities probably smoke cigarettes at least once a month The perceived popularity among successful/elite elements of society scale was formed by summing the items and taking the mean as the score The internal consistency score (Cronbach’s a) was 55 The low internal consistency here reflects that these items were not uniform in measuring the same constructs (eg, perceptions of wealthy people and favorite celebrities appeared to not be closely related) Perceived prevalence items asked students to estimate what percent of different groups of people (secondary students, college students, and adults) had smoked a cigarette in the past 30 days The response options for these items were on an 11-point scale, from through 100 in 10-point increments The items referring to secondary students measured perceived prevalence of 10th, 11th, and 12th grade students The perceived prevalence of smoking scale was formed by summing the five perceived prevalence items and taking the mean as the score The internal consistency score (Cronbach’s a) was 91 Current smoking was measured and operationalized as having smoked cigarettes at least once in the past 30 days Ever smoked cigarettes was measured by asking students “have you ever tried or experimented with cigarette smoking, even a few puffs?” Susceptibility to future smoking was assessed with the measurement technique developed by Pierce et al27 which assesses “nonsusceptibility” as answering “definitely no” to the following items: (a) Do you think that you will smoke a cigarette soon? (b) Do you think you will smoke a cigarette in the next year? (c) If one of your best friends were to offer you a cigarette, would you smoke it? Other smoking variables that were assessed were as follows: age of first smoking, ever experimented with smoking, and number of cigarettes smoked per day in the past 30 days Demographic information included gender, grade level, and age The survey also asked students “In general what grade you get in school?” and students responded on a scale of to (with being the highest) The survey instrument was translated from English into Vietnamese and back translated into English to ensure equivalence of the English and Vietnamese survey versions Before survey administration, the instrument was pilot tested in a secondary school in Hanoi to ensure that students understood survey items without difficulty Data Analysis The prevalence of current smoking, ever tried smoking, susceptibility to future smoking, and other smoking behavior was calculated and c2 tests were used to test for differences between boys and girls on these variables To determine if the items representing normative beliefs about smoking were part of one, or more than one underlying concept(s) the factor analysis procedure used by Primack et al19 was followed as a method of analysis in the study Thus, iterative principal components analysis with varimax rotation was used to determine the underlying factor structure of the normative belief items To examine the relationship of normative beliefs (perceived parent/peer disapproval, perceived prevalence of smoking, popularity among successful/elite) and smoking behavior (current 72 Asia-Pacific Journal of Public Health 24(1) smoking, ever tried smoking, and susceptibility to smoking in the future), bivariate analysis was first conducted Following bivariate analysis, logistic regression was used to determine the relationship of the normative belief scales with the following as dependent dichotomous variables: (a) current smoking, (b) ever tried smoking, and (c) susceptibility to smoking in the future The multivariate logistic regression models that examined the association of normative beliefs with these dependent dichotomous variables (current smoking, ever tried smoking, susceptibility to future smoking) also included gender, grade level, age, and self-reported performance in school as predictor variables Thus, gender, grade level, age, and self-reported performance in school were controlled as potential confounders in the logistic regression models In analyses involving susceptibility to smoking as a variable, only nonsmokers were included because this construct was developed and validated among nonsmoking adolescents by Pierce et al.27 To test for moderating effects of gender, grade level and school, interaction terms were formed by creating crossproducts with the three social normative beliefs (perceived parent/peer disapproval, perceived prevalence of smoking, popularity among successful/elite) and included in the logistic regression analyses predicting current smoking, ever tried smoking, and susceptibility to smoking The Wald c2 statistics were evaluated to determine if there were significant interaction effects with gender, grade level, and school Results The prevalence of smoking behavior and susceptibility to future smoking by grade level and gender for both samples is presented in Table Principal components analysis on the 12 normative belief items revealed a clear 3-factor solution as independent dimensions/factors of normative beliefs in this sample of Vietnamese students (Table 2) The eigenvalue was 3.87 for the factor associated with the perceived prevalence of smoking items, 2.29 for the factor associated with the disapproval of smoking by parents/peers items, and 1.80 for the factor associated with the popularity of smoking among successful/elite items These factors, explained 32%, 19%, and 14% of the variance, respectively Table also provides the mean scores on each item In the fully adjusted logistic regression models (see Table 3), parent/peer disapproval was independently associated with a higher risk of current smoking, ever trying smoking, and susceptibility to smoking Perceived prevalence of smoking was associated with current smoking and susceptibility to smoking among Vietnamese students and appeared to be marginally associated with ever tried smoking even though it did not reach statistical significance Perceived popularity among successful elite was not associated with any of the smoking behavior variables The associations of the demographic variables with smoking behaviors are also included in Table Male gender was consistently associated with the smoking behaviors, showing that males were at significantly higher risk of smoking and susceptibility to future smoking than females As a result of association of smoking with parental/peer disapproval, we conducted follow-up logistic regression analyses to determine whether parental or peer disapproval of smoking had a stronger protective effect against the smoking behaviors In predicting current smoking and controlling for the same covariates as in the previous analyses, the odds ratios were 0.62 (0.41, 0.93) for disapproval of friends, 0.74 (0.50, 1.08) for disapproval of most people my age, and 0.76 (0.56, 1.03) for disapproval of parents A similar pattern showing the highest protection from disapproval of friends was found for ever tried smoking The odds ratios for susceptibility to smoking in the future were 0.60 (0.48, 0.74) for disapproval of most people my age, 0.77 (0.61, 0.97) for disapproval of friends, and 0.83 (0.69, 0.99) for disapproval of parents Given the lack of association of smoking with the perception of smoking among successful/ elite members of society and also the low internal consistency of this scale, we conducted additional logistic regression analyses to determine whether the individual items were predictive of 73 Page et al Table Smoking Behavior and Susceptibility to Future Smoking Percentage (n) Significance Boys Girls c (df) Current smoking 10th Grade 6.9 (22) 1.2 (3) 11th Grade 9.4 (24) 1.2 (3) 12th Grade 11.8 (33) 1.5 (6) Total 9.2 (79) 1.3 (13) Susceptibility to future smokinga 10th Grade 44.1 (141) 8.3 (21) 11th Grade 47.8 (123) 12.3 (42) 12th Grade 51.2 (144) 11.0 (45) Total 47.5 (408) 10.7 (108) Ever tried or experimented with cigarette smoking 10th Grade 29.1 (93) 6.7 (17) 11th Grade 22.6 (58) 7.0 (21) 12th Grade 33.6 (94) 6.1 (25) Total 28.6 (245) 6.6 (66) Daily smoking 10th Grade 1.6 (5) 0.8 (2) 11th Grade 3.1 (8) 0.0 (0) 12th Grade 4.6 (13) 0.2 (1) Total 3.0 (26) 0.3 (3) Smoked a cigarette at 12 years of age or younger 10th Grade 10.3 (33) 0.8 (2) 11th Grade 9.3 (24) 2.1 (7) 12th Grade 18.9 (53) 2.7 (11) Total 12.8 (110) 2.0 (20) P 11.01 (1) 22.08 (1) 33.25 (1) 62.04 (1) .0009b