Predictors of cigarette smoking among adolescents in rural Zambia : results from a cross sectional study from Chongwe district

Introduction: Cigarette smoking is a leading cause of morbidity and mortality globally. There are limited data on the prevalence of and factors associated with smoking among in-school adolescents in developing countries. Objectives: To estimate prevalence of those who have smoked cigarettes and to identify associated socio-demographic factors among adolescents in Chongwe district, Lusaka Province, Zambia. Methods: Data used was from the Zambia Global Youth Tobacco Survey, which was conducted using standardized methodology among in-school adolescents in 2002. Data were analyzed to assess if selected socio-demographic variables were associated with having smoked cigarettes. Bivariate and multivariate logistic regression analysis was used to assess the associations. Results: A history of having smoked cigarettes ranged from 20.5% among 15 year olds to 37.2% among males younger than 12 years old. In females, 20.7% of 13 year olds and 37.7% of those less than 12 years old had smoked. Parental smoking, friends smoking, a lack of perception that smoking was harmful, and exposure to pro-tobacco advertisements were associated with having smoked cigarettes. Adolescents who had smoked cigarettes were more likely to allow others smoke in their presence. Conclusions: Many adolescents in rural Luangwa, Zambia had tried cigarette smoking. The identification of predictors for smoking should guide the design and implementation of programs aimed to prevent initiation and maintenance of tobacco use among adolescents.


Introduction
Cigarette smoking is the single most common cause of prevantable morbidity and mortality globally.It is estimated that by 2020, tobacco-related deaths will be the leading cause of deaths in developing countries 1,2 .
While smoking is an important disease risk factor on its own, many adolsecents who smoke also have unhealthy lifestyle clusters such as smoking, alcohol use and illicit drug use 3 .Smoking may, therefore, be a marker of other harmful adolescents behaviours.
Research into adolescent cigarette smoking has mostly been conducted with current smoking status as the main outcome.However, the history of previous smoking is an important predictor of current or future smoking 4 .Also, studies into adolescent cigarette smoking have generally reported on the prevalence of smoking without consideration of relevant potential associations with socio-demographic factors.However, the identification of such factors can best inform the design and implementation of public health interventions aimed to reduce adolescent tobacco use.
We report on a study aimed to estimate the prevalence of having smoked cigarettes, and its association with relevant socio-demographic characteristics.

Methods
The data were obtained in a cross-sectional school-

Distribution of having smoked cigarettes by socio-economic variables
The prevalence of a history of having smoked cigarettes -even a single puff -was assessed.The results are presented (Table 1).Overall, the proportion of adolescents who reported having smoked was 27.0%.
(26.8-27.2).Generally all age groups had a prevelance exceeding 20%.However, the extremes of the age group had the highest history of having smoked.A similar dsitribution was observed for class grades where the lowest and highest grades had higher prevalence than the middle grade.Adolescents whose parents were smokers or whose friends were smokers were themsleves more likely to have smoked.

Attitudes towards smoking
It was also aimed to assess attitudes towards smoking.
In general, study participants had negative attitudes towards tobacco or smoking.There were, however, exceptions: for instance, the finding that favouring baning smoking in public and the perception that environmental exposure was harmful were associated with having smoked.The other results are presented (Table 2).

Tobacco advertisements and smoking
It was also aimed to determine if having smoked was associated with exposure to different forms of tobacco advertisements, owning a cigarette-logo branded item, and exposure to pro-tobacco advertisements at sports or social events.Exposure to both forms of advertisements were associated with having been a smoker.Detailed results are presented (Table 3).

Discussion
Our study found that the prevalence of having smoked cigarettes among in-school adolescents in Chongwe district, Zambia was 27%.This prevalence is lower than the 31% reported by Mpabulungi and Muula for Arua, rural Uganda 8 , and the reason for this is unknown.One possible explanation is that Arua's economy largely depends on tobacco farming, while Chongwe's does not, and so there may be a greater tolerance of tobacco use in Arua.Other unknown country-level factors may account for the differences between the two districts (eg in the Greek GYTS, Kyrlesi et al. 9 reported a 32.2% prevalence of having smoked among 13-15 year-old adolescents).While the present study revealed no gender differences, another study of Zambian adults identified a higher prevalence among males 10 .Our study identified that having a parent who smoked was associated with the adolescent having smoked.
Ornelas et al. 11 , in their US study on parental influence on adolescent physical activity, found that parent-child communication and family cohesiveness were important influencing factors.Although that study found that parental monitoring was not a good predictor of adolescent physical activity, Arredondo et al. 12 reported that parental monitoring was a good predictor of adolescent healthy eating and physical actovity.Factors involved in parents influencing adolescents' smoking may include permissiveness, the availability of cigarettes in the home and role modeling, and this should be investigated further.

Thomas et al.'s systematic review of family-based
tobacco-control programs aimed at preventing adolescent smoking 13 , suggested that well-conducted programs were effective.School-based programs may also be effective 14 .
In our study, adolescents who perceived tobacco smoke to be harmful to their health were less likely to have smoked than those who were not concerned about the harmful effects of tobacco.This was also reported by Rodriguez et al 15 .This suggests an opportunity to change adolescents' beliefs about smoking, by using the media to reduce smoking behavior.
It has been established that tobacco promotion directed at adolescents enocourages the uptake and maintenance of smoking 16,17 some of the health centers are located over 40 km from main roads7 .Data analysisData were analysed using SPSS v 14.0 (SPSS Inc; Chicago, IL, USA).It was aimed to assess the prevalence of having smoked cigarettes, stratified by sex.Bivariate logistic regression analyses were conducted to obtain an unadjusted odds ratio between having smoked and explanatory variables of interest (Model 1).Model 2 was multivariate analysis with the explanatory variable of interest as the main exposure controlling for the other variables.Ethical considerationsPermission to conduct this study was obtained from the Zambian Ministry of Education.Parents and guardians were informed of the study.Parents who did not want their children to participate were asked to inform their children not to particiapte.All students were informed that they were free not to participate or not to answer any questions on the questionnaire.The questionnaire was completed in class time.Students who did not want to particiapte remained in class but could submit an uncompleted questionnaire, because excluding them from the class was potentially stigmatizing.To preserve indidividual confidentiality, the questionnaires were self-completed anonymously by the student.School teachers were not in the classrooms at the time of questionnaire completion.

Table I :
Demographic factors associated with having tried cigarette smoking among school going © S Siziya, E Rudatsikira, AS Muula, PRT Ntata, 2007.A licence to publish this material has been given to ARHEN http://www.rrh.org.au 5

Table 2 :
Attitudes about smoking, and factors associated with having tried cigarette smoking among school going adolescents in Chongwe district, Zambia Siziya, E Rudatsikira, AS Muula, PRT Ntata, 2007.A licence to publish this material has been given to ARHEN http://www.rrh.org.au6

Table 3 :
Associations between advertisements and having tried cigarette smoking among school-going adolescents in Chongwe district, Zambia