Prevalence, patterns and correlates of alcohol consumption and its' association with tobacco smoking among Sri Lankan adults: a cross-sectional study.
Katulanda P., Ranasinghe C., Rathnapala A., Karunaratne N., Sheriff R., Matthews D.
BACKGROUND: Most studies on alcohol consumption carried out in Sri Lanka are limited to single/few provinces in the island. The objective of this study was to determine the prevalence, patterns and correlates of alcohol consumption among a larger sample of adults in Sri Lanka. METHODS: This cross-sectional study was conducted in seven of all nine provinces in Sri Lanka, between 2005 and 2006. A nationally representative sample of 5000 adults aged ≥18 years was selected using multi-stage random cluster sampling. Data of 4532 participants were collected using an interviewer administered questionnaire. Data analysis included chi-squared test, multiple logistic regression analysis and Spearman correlation using Stata/SE 10.0 (StataCorp LP., Texas, USA) software package. RESULTS: Males were 40%; mean age was 46.1 years (±15.1). The overall, urban and rural prevalence (95% CI) of current drinking was 23.7% (21.7 - 25.7), 29.5% (25.7 - 33.3) and 22.2% (19.8 - 24.7) respectively (p = 0.001). Current (M: 48.1%, F: 1.2%, p < 0.0001) and former (M: 21.4%, F: 0.7%, p < 0.0001) drinking was much higher in males. The highest prevalence of drinking in males (58.9%) and females (2.2%) was in the 30 - 39 and <20 year age groups respectively. Lowest prevalence in men (24.6%) and women (0%) was in the >70 years age-group. Hazardous drinking was seen in 5.2% of men and 0.02% of women. Male sex, urban living and current smoking correlated with both current and hazardous drinking. Lower level of education, and age >70 years positively correlated with hazardous drinking. CONCLUSIONS: Alcohol is predominantly a problem in Sri Lankan males. In males, both current and hazardous drinking positively correlated with urban living, white collar occupation, Burgher ethnicity and current smoking. Hazardous drinking positively correlated with lower level of education and older age. The data shown here are useful in planning interventions simultaneously targeting alcohol and tobacco.