To estimate if there is a non-linear association between varying levels of alcohol use and poor mental health (depressive and post-traumatic stress disorder (PTSD) symptoms) in university students from low-, middle- and high-income countries. Using anonymous questionnaires, data were collected from 19 238 undergraduate university students (mean age 20.8; standard deviation (SD) 2.8) from 27 universities in 26 countries across Asia, Africa and the Americas. Alcohol use was assessed in terms of number of drinks in the past 2 weeks and number of drinks per episode, and measures of depression and PTSD symptoms were administered. The proportion of students with elevated depression scores was 12.3%, 16.9%, and 11.5% for non-drinkers, moderate drinkers, and heavy drinkers, respectively, while the proportion of students with high PTSD symptoms was 20.6%, 20.4% and 23.1% for non-drinkers, moderate drinkers, and heavy drinkers, respectively. Logistic regression found that non-drinkers and heavy drinkers had a lower odds than moderate drinkers to have severe depression, after adjusting for sociodemographic variables, social support and subjective health status. Further, heavy, more frequent drinkers and more frequent binge drinkers had a higher odds to have elevated PTSD symptoms than moderate and non-drinkers, after adjusting for sociodemographic variables, social support and subjective health status. The results suggest a reverse U-shaped association between recent alcohol use volume and frequency and depressive symptoms (unlike that previously identified), and a J-shaped association between binge drinking frequency and depressive symptoms and alcohol use and PTSD symptoms.
Human Sciences Research Council (Sat,) studied this question.