This report provides an overview of Demographic and Health Surveys (DHS) survey questions related to alcohol consumption and an analysis of outcomes from these questions in DHS surveys. It first examines the questions on alcohol use and the response categories in all surveys that included questions about ever, typical, or current alcohol consumption. The inventory found that alcohol questions were included in 65 surveys in 42 developing countries between 1987 and 2014, and 19 different questions related to alcohol consumption were identified. The most common question asked if the respondent had ever drunk alcohol. Responses to this question were used to calculate the prevalence of ever-drinking in 36 surveys. Prevalence varied substantially within and across regions and between men and women. The study then uses multivariable logistic regression to examine the association between alcohol use and unsafe sex in eight countries in sub- Saharan Africa known to have high HIV prevalence. The regression analysis of the global association between self-reported drinking status and unsafe sex was not found to be significant except among men in Zambia, where the odds of engaging in unsafe sex for a current drinker were lower than for a non-drinker (opposite the expected outcome). At the event level, alcohol consumption at last sex was an inconsistent predictor of unsafe sex. For both men and women, the strongest predictor of engaging in unsafe sex was age of respondent—higher odds for those 15-24 compared with those 35 and older. For women, lifetime number of sexual partners also strongly predicted unsafe sex. The authors suggest using validated scales to assess harmful drinking.
Record created by