Tobacco use and its predictors among Ethiopian adults: A further analysis of Ethiopian NCD STEPS survey-2015

Background: Consuming any form of tobacco is one of the leading causes of preventable morbidity and mortality. Tobacco smoking has been identified as one of the major risk factors for NCDs, including cardiovascular, chronic respiratory diseases, and different cancers. Although there is national information on magnitude of tobacco use, to date there is limited nationally representative data on factors associated with tobacco use. The aim of this study is to assess the distribution and predicators of tobacco use in Ethiopian adult population between 15 -59. Objectives: The main aim of this study was to assess the prevalence of tobacco use and its predictor in Ethiopia. Methods: A cross-sectional population based study design was employed among population age from 15- 69 years. A stratified, three-stage cluster sampling was used to identify the study subjects. Households in each cluster were selected using simple random sampling method. The sampling frame was based on the population and housing census conducted for Ethiopia in 2007. Data was collected using WHO NCD STEPS questionnaire; current tobacco use of any type was taken as the dependent variable. Five hundred thirteen enumeration areas (EAs) as primary sampling units (PSUs) (404 rural and 109 urban) were selected with probability proportionate to size, followed by selection of households as a secondary sampling units (SSUs). A total of 10,260 households were selected from the 513 EAs (20 households per EA). Eligible individuals were selected from households using Kish method (a pre-assigned table of random numbers to find the person to be interviewed). Descriptive statistics using frequency table, mean, median, interquartile range and standard deviations were computed. Step wise logistic regression was used to analyse the predictors of tobacco use. An Estimator of 95% confidence interval was used both for computing descriptive statistics as well testing associations using logistic regression. Results: The prevalence of tobacco use (all tobacco products) was 4.2%. The mean age (± SD) of starting tobacco use was 21(7) years. The mean frequency of tobacco use was 2 times per day. Hierarchical Logistic regression analysis revealed that participants in age groups 30-44 years, and 60-69 years were less likely to use any tobacco type compared to younger age group of15-29 years. Heavy episodic drinking, AOR 2.46 [95% CI= 1.4 – 4.5], and khat chewing, AOR 4.71[95%= 2.26 – 9.8], were independently associated wit