Jornal de Comunicações de Cuidados de Saúde Acesso livre

Abstrato

Differential Burden and Determinants of Tobacco Smoking: Population-Based Observations from the Zambia Demographic and Health Survey (2002 and 2007)

Pawel Olowski, Charles Michelo

Background: Tobacco smoking has been said to be on the increase in sub-Saharan countries which could potentially result in higher rates of associated non-communicable diseases (NCDs). We investigated the prevalence and factors associated with tobacco smoking in Zambian general population, and how they changed over time. Method: Data stem from the Zambia Demographic and Health Survey conducted in 2002 (n=9803) and 2007 (n=13646). Extracted data on men aged 15-59 and women aged 15-49 was analyzed using a weighted analysis in STATA 12. Descriptive statistics involved investigation of demographic and socio-economic features of the study population. Predictors of tobacco smoking were examined using bivariate and multivariate stepwise binary logistic regression stratified by sex. Tobacco smoking prevalence and its associated factors were compared between 2002 and 2007 to assess the change. Overall non-response rates in these surveys were 5% and 6% respectively. Results: Of all the respondents in 2002 (n=9803) 22% were men and 78% were women, whereas in 2007 (n=13646) 48% were men and 52% were women. Tobacco smoking prevalence decreased by about 2% in 2007 for both men (26.4% vs. 24.2%, P=0.038) and women (2.7% vs. 0.8%, P<0.001). Prevalence of smoking cigarettes in rural settings increased from 12% to 26% (P<0.001) among men, and from 0.4% to 1.1% (P<0.001) among women while it did not change in urban areas. Older ages, drinking alcohol, and lower education levels were common factors associated with higher odds of tobacco smoking in both 2002 and 2007 survey rounds. Conclusion: We find contrasting rural urban smoking burden differentials contributing to the unchanging smoking burden overall in this population. The concentration of this burden in rural and less educated groups suggests past limitations to health promotion initiatives. This calls for a reformation in behavior information targeting these most at risk groups.