Modeling and Predicting Prevalence of Tobacco Use in Bangladesh

Abstract

Tobacco consumption is one of the major preventable causes of death and disability globally. Bangladesh is one of the largest tobacco consuming countries in the world. The use of tobacco is high among male population in Bangladesh. Also the use of tobacco is increasing. Therefore this study aims to explore the prevalence, patterns and determinants of tobacco use among adults. This study used data from Global Adult Tobacco Survey (GATS), 2010 that covered 9,629 Bangladeshi aged 15 years or above. The survey was based on a three-stage stratified cluster sample of household. Information of a total of 9629 adults has been analyzed in the present study. Pattern and prevalence of tobacco consumption in Bangladesh are primarily presented in the form of frequency distributions and in prevalence rates. This study uses binary logistic regression model and multilevel logistic regression model for analyzing data to quantify the objectives of the study. Prevalence of adults tobacco user was 44.05% in Bangladesh with 65.11% in male and 34.89% in female. Common significant predictors include sex, age, educational level and wealth Index. Male and older had a higher tendency to use tobacco products. Current tobacco smoking was significantly higher among male (p-value<0.001 and OR=44.17) than female. Females were more likely to use smokeless tobacco than males (odds ratio, OR=1.72). Adults with no education were more likely to use tobacco products in Bangladesh compared to others with tertiary education. Adults with the poorest wealth status were more likely to consume tobacco products in Bangladesh compared to those from richest wealth index. It applies a multilevel (two level) logistic regression analysis to draw valid conclusions about the effects of the selected determinants on tobacco consumption using GATS-2010 data which is a multistage stratified cluster data. Instead of single level logistic model, multilevel logistic regression model has been utilized since the data follow a hierarchical structure. Also the comparison between single and multilevel model has been done to investigate the necessity of multilevel effects. The findings suggest that sex, age, level of education and wealth index have significant multilevel effects on tobacco consumption. Manufactured cigarettes and betel quid with zarda are the most usable tobacco products in Bangladesh. Intervention to reduce or stop tobacco using should be directed towards the poor, older and people with lower education and women should be targeted for prevention of the use of tobacco. All the significant variables should be considered for developing suitable policies to reduce the consequences of tobacco use in Bangladesh.